october 2000 issue 2 3 contributors 4 6 ... - forex-warez.comforex-warez.com/trading books/magazine...

OCTOBER 2000 ISSUE

2 Editor’s Note

3 Contributors

4 Web Watch

6 New Products

7 Hardware, Software and Communications: Tech frontiersWith promises of voice-activated software, increased mobility and greater access to global markets, emerging trading technology proves the once far-out future isn’t far off.

10 Product Review:Online Trading Academy’s Fundamentals of Direct-Access Trading

12 What you see is (not) what you getUnderstanding market maker sleights of hand on the Level II screen.

16 Opening day opportunitiesHow to use an intraday pullback to catch a ride on a hot IPO.

19 More bang for your buck: Patterns within patternsCreate trade opportunities with increased reward and decreased risk by trading patterns within patterns.

23 Trading tests and retracements with Japanese candlesticksHow to use simple Japanese candlestick patterns at test and retracement points.

28 Moving beyond the closing priceHere’s an advanced technique that uses price changes rather than specific price levels to calculate indicators that can help project likely future price ranges.

33 The Face of Trading:The overnight option

35 Trading Psychology: Conquering trading biasesTrading coach Van K. Tharp discusses the psychological basis of successful trading.

38 Less is moreBy keeping track of the momentum of your trades you will get a good feel for what your next move should be — taking profits, or even adding to an already profitable position.

42 Indicator Insight: Momentum and rate of changeOverview of momentum and rate of change (ROC)

45 Trading through turmoilHistory tells us that periods of market turmoil are unavoidable, even if they’re also unpredictable. Understanding how to use instruments like the index tracking stocks can help keep you afloat during such times.

49 Profiting from intermarket analysis Individual trades can’t occur in an information vacuum. Analyzing the interaction of market indices can help you find the best stocks in the best sectors to trade at any given time.

54 Trading System Lab: RS system No. 1

ACTIVE TRADER • October 2000 • www.activetradermag.com 1

2 www.activetradermag.com • October 2000 • ACTIVE TRADER

M ore tools, more speed, greater reach.Never has there been so much technology devoted to giving

traders greater power over their trades — how they get anduse the information that goes into making a trading decision,how they execute their trades and how much they pay forthem.

But with that power comes responsibility. Much tradingtechnology is the proverbial double-edged sword: Its potentialrewards can only be reaped if you have the time and patienceto learn its nuances and accept its realities.

You can go out and buy a set of top-of-the-line tools or a sports car, but that does-n’t automatically make you a master car-penter or a Formula One driver. Similarly,a trader armed with real-time data and adirect-access brokerage does not become aseasoned market maker overnight. In fact,he or she can’t become a market maker —period — because, despite the technologi-cal advances retail traders now enjoy, theexisting Nasdaq order execution rules stillheavily favor professional market makers.

We’ve commented before on the mythof the level playing field (see “Level IIquotes: Decoding supply and demand,”Active Trader, June, p. 30). Simply, marketmakers are allowed to do things we can’t,such as watch order flow come in whilethey decide whether or not to take a particular trade (or onlyfill part of it). Although developments such as SuperSoes andSuperMontage (see “Inside the Market,” Active Tr a d e r,September, p. 15) may eliminate some of these advantages (ifand when they come to pass), for now, the game is what it is.If you want to play, you have to pay.

That it’s still a market maker’s game is partially the subjectof “What you see is (not) what you get” on p. 36. This storyunderscores the reality that direct-access, Level II trading is abattlefield that must be tread with open eyes and instincts justas quick to flight as fight, depending on the situation. If you’renot familiar with the games market makers play — biddingwhen they need to sell, offering when they want to buy, hidingsize and shifting orders — you can quickly find yourself on thelosing end of the trading equation. Careful study and lengthypreparation are necessary to be able to recognize the patternsthat provide a clearer picture of order flow.

Also in the Trading Strategies section is a look at a short-term IPO strategy designed to isolate an intraday pullback andcatch stocks as they’re taking off. The “hot IPO” market might

have cooled off lately, but there are still ways to take advantageof profitable short-term moves in new stocks.

It’s not often you can have your trading cake and eat it, too,but in “More bang for your buck: patterns within patterns,” (p.48) we look at how you can structure trades with very short-term risk and longer-term profit potential by looking for con-solidations within consolidations. Rounding out the Strategiessection is an article by Teresa Lo on how to combine Western

test and retracement analysis withJapanese candlestick reversal patterns.

Also this month, we kick off a new sec-tion, “Advanced Strategies,” dedicated toh i g h e r-level trading and analysis tech-niques. Senior editor Thomas Stridsmanaddresses two key weaknesses of manycommon technical indicators and illus-trates ways to build more reliable, statisti-cally sound short-term trading tools.

In this month’s Hardware, Software &Communications story (“Tech frontier,” p.30), CyberCorp chairman Philip Berberpoints out the hurdles to tomorrow’s elec-tronic, global, 24-hour direct-trading envi-ronment have more to do with regulationthan technology. However, it is technologythat is forcing proposed changes in trad-

ing regulations and how exchanges do business. In our Inside the Market section, associate editor Jeff

Ponczak takes a look at some long-standing rules — includingthe short-sale (uptick) rule, and the selective disclosure ofimportant company/financial news to preferred clients andanalysts — that are coming into question or falling by the way-side.

In today’s trading world, the challenge doesn’t seem to be somuch developing the necessary technology to empowertraders, but developing a market and regulatory structure thatcan respond quickly enough to the sweeping changes technol-ogy is forcing on the trading industry.

Wielding the tools of trading

Mark Etzkorn, Editor-in-chief

Trading technology

is the proverbial

double-edgedsword.

EDITOR’SNote


THIS MONTH’S Contributors

▼ Dewey E. Burchett is president and senior editor of StockLogix.Com

(www.StockLogix.com), located in Shreveport, La. He is a registered

investment advisor and writes daily market commentary that includes both

technical and fundamental analysis geared for the active trader/investor. Prior

to launching StockLogix.Com (formerly Online Daytraders.com) in March 1999,

Burchett managed trading accounts for individuals and trained traders at a day

trading brokerage.

▼ Steve Wendlandt has more than 15 years of trading experience in the stock,

futures and options markets. He is currently the chairman of Sequoia Capital

Management Inc. and a full-time trader with Online Investment Services in

Houston. He was also a former Commodity Trading Advisor specializing in

mechanical systems trading. Wendlandt primarily focuses his time researching

and trading U.S. equities and options. He can be reached at (713) 822-5855 or

[email protected].

▼ Teresa Lo worked in the stock brokerage business for more than a decade prior

to retirement from the industry in 1998. She is a technical trader and uses

simple, classic techniques to analyze the market. She holds a BAin Economics

and Psychology from the University of British Columbia. Lo is co-founder and

conscience of Intelligent Speculator, www.intelligentspeculator.com, an original

non-commercial Web site focusing on technical trading, risk and money

management. It features extensive resources for those who would like to

improve their trading performance. She can be reached at

[email protected].

▼ Michael Chalek is a commodity trading advisor and market consultant.

In addition, he manages a private pool at One World Capital Management.

He is president of Universal Technical Systems Inc., a firm dedicated to the

research and development of innovative market techniques. He can be reached

via his Web site at www.tradefutures.com or e-mail at [email protected].

▼Michael A. Mermer is a registered commodity trading advisor and president

of Traders Software Company, developer of the ETS System and publisher of the

ETS Real-time Signals (www.traderssoftware.com and www.stock-trading.com).

He can be reached at (800) 637-6529 or [email protected].

▼ John Saleeby is a Chicago-based trader who was featured in the May

Active Trader Interview (see “John Saleeby: Mastering the trading arcade”

on p. 70 of that issue). During the course of the interview, Saleeby made the

stock market calls (and trades) that are the subject of his article in this

month’s Big Picture section (“Profiting from intermarket analysis,”

p. 96). He can be reached at [email protected].

▼ Kiara Ashanti is a Florida-based writer, editor and private trader. He has been

in the financial industry for the last five years, and trading the last two.

As a freelance writer, he has written for BET.com, the Tampa Tribune, and Profit

Magazine. Currently, he is working on a book about day trading.


WEBWatchBig screen

I n the past few years, there’s been no dearth of Web sitesthat allow traders to use and customize charts, and newsand re s e a rch sites have been around almost as long as theWeb itself. A newer and expanding “genre” of trading-

related Web sites is the screening site, which allows traders tofind stocks that meet specific technical and fundamental criteria.

One of the latest entries into the field is MarketScreen(www.marketscreen.com). On the home page, in addition toselected news stories from six different organizations, is one ofMarketScreen’s pre-selected stock screening lists. Using a drop-down menu, you can sort stocks based on 21 different screen-ing categories, ranging from Bollinger Bands to MovingAverage Convergence/Divergence and MACD Histogram toMarket Overview to Weighted Moving Average Crossover.

The subcategories for each screen are different. For example,in the Relative Price category, there are 10 different criteria suchas new 52-week high, 40 percent off 52-week low and 80 per-cent off 52-week high. In the Market Overview category, thereare 17 different sub-sections, from trending down to stochasticbullish reversal to MACD crosses below zero (there’s an exten-sive education section on the site that provides an explanationof all these different terms; often you can click on the term anda separate box with a definition will appear). Unfortunately,

searches are not real-time — MarketScreen’s biggest drawback.All screens are based on a stock’s most recent closing price (i.e.,Tuesday’s screens will be based on Monday’s close).

When showing screening results, MarketScreen separatesstocks into Nasdaq, NYSE and American Stock Exchange, theninto Large Cap, Mid Cap, Small Cap and Micro Cap, showingthe number of stocks in each category. On the day we reviewedthe site, for example, there were 29 Nasdaq Mid-Cap stockswhose 7-day moving averages crossed their 13-day movingaverages. Clicking on the number gives you a list of all thestocks along with their previous closing price, volume and per-centage from 52-week high and low. You can get a profileand/or quote of all the stocks (both courtesy of SiliconInvestor), or you can access a chart, which, if you’re a regis-tered member, is customizable (registration is free).

M a r k e t S c reen allows you to configure your own screens andcombine diff e rent screening criteria. You must be a re g i s t e re duser to access customized screens. You can mix and match 16d i ff e rent fundamental indicators (such as market cap, P/E ratio,earnings) and 20 diff e rent technical indicators (such as EMAc ro s s o v e r, relative price, candlesticks). Unfortunately, you canonly mix technicals with technicals and fundamentals with fun-damentals; you cannot cross-mix the indicators.

MARKETSCREEN.COM

MarketScreen.com allows you to search for stocks using a wide variety of criteria.


On the chart

T he numerous interactive charting sites available onthe Web won’t do you any good if you don’t have abasic understanding of charts (and the indicators

that go with them) in the first place. ChartHelp(www.charthelp.com), which only began in July, is just whatit sounds like — an educational site that provides a goodexplanation of differentchart patterns and indi-cators. Best of all, it’sfree (registration isrequired).

A c t u a l l y, the sitegives minimal informa-tion on the four basictypes of price charts(bar, line, point and fig-ure, and candlestick),but it more than makesup for it with the detailit has on patterns andindicators. For starters,you can download (ineither Adobe Acrobat orWord) the formulae for18 different indicators,from simple and expo-nential moving average to the swing index.

The bare minimum needed to decipher price charts is anunderstanding of basic patterns, and ChartHelp does a goodjob of explaining trendlines, breakouts, triangles, and topsand bottoms. Descriptions and rules of interpretation ofeach specific type are given (i.e., support and resistance,ascending and descending triangles, head-and-shoulders)with visual help provided by price chart examples.

In-depth chart analysis, though, requires knowledge ofnumerous indicators, and ChartHelp lists 50. However,because of the newness of the site, it doesn’t give a defini-

tion or analysis of all of them. That’s too bad, because whenChartHelp does explain an indicator, it does a thorough jobof it, again using visual examples from charts to make it eas-ier to understand.

ChartHelp eventually plans on having a definition andexplanation for all the indicators, and when they do, there’s

an on-site charting option so you can apply what you’ve justlearned. There’s also a fairly extensive technical glossary(you can have it open in another window so you always haveaccess to it), and ChartHelp will send you a bi-monthlynewsletter. (The first newsletter dealt with trendlines andmoving averages; the second with the moving average con-vergence/divergence.)

So if you’re new to the trading game or just looking tobrush up on your technical analysis basics, chart a course toChartHelp next time you’re surfing the Web.Ý

CHARTHELP.COM

ChartHelp.com provides information on charts and indicators.

You can, for example, search for stocks that have gapped upm o re than 5 percent on bullish On Balance Volume. Or, stocks inthe financial industry priced between 50 and 75 that closed at a52-week high. All customized searches can be saved and usedagain. However, you cannot adjust the parameters of individualindicators (the number of days used to calculate a moving aver-age or stochastic, and so on). The site shows screening results forseveral indicator-length values (for example, 20 diff e rent mov-ing average combinations), but this still isn’t the same as beingable to customize your own indicators to use in a screen.

Another benefit for registered users is the stock alert feature.Users can be alerted daily via e-mail as new stocks meet theirscreening criteria.

M a r k e t S c reen also offers an Event Profile. This allows you totype in the name of a stock and receive all the categories it fits into

for any day from the most recent close to the close five days ago.For example, when we typed in ORCL (Oracle), MarketScre e ntold us on that particular day Oracle met 20 diff e rent criteria,such as it closed above its 7-day weighted moving average andits OBV was bullish. However, when we checked the site overseveral days in mid-August, it had a bad habit of listing everystock as 20, 40, 60 and 80 percent off its 52-week low or 52-weekhigh — even the ones that did not come close to meeting this cri-teria. And, when you clicked on these terms to get a definition, allthat came up was an error message.

Despite the lack of real-time data, MarketScreen is a usefulsearch tool as its many possible permutations allow traders toget quite specific in their screens. That, coupled with the chart-ing and e-mail delivery options, makes MarketScreen a handytool for traders.Ý


NEW Products▼ Terra Nova Trading (www.terranovatrading.com) now offers its

customers business-class DSL access. This comes as a result of a deal

between the online brokerage firm and broadband service pro v i d e r s

Big Net and Covad Communications. Terra Nova clients who execute

m o re than 50 trades per month will receive the DSL connection fre e .

Less active traders can get the service at a discount.

▼ On-Site Trading Inc. (www.onsitetrading.com) has partnered with

trading and investment software developer Neovest Inc.

(www.neovest.com) to add Neovest’s FirstAlert trading software to its

system. FirstAlert filters stocks according to criteria set by the trader

and alerts the user to actionable opportunities as they occur in real

time. The On-Site Trading version of the FirstAlert system will include

the company’s proprietary S.M.A.R.T. Trading Technology, which

searches all market makers and ECNs for the best price.

▼ SwingWire (www.swingwire.com) is an online investment “com-

munity” where investors and traders can share real-time market

ideas, stay updated on market conditions and movements, and learn

from experienced market players. Launched in late August, the site

allows members to test their stock-picking skills or new trading meth-

ods in real time, access chat and message boards, and view, in depth,

portfolios of the SwingWire CyberAnalysts. Visit the Web site for

more information on joining or on becoming a CyberAnalyst.

▼ Traders using MarketXT can now get after-hours equity market

news and analysis from MidnightTrader.com, as a result of a recent

partnership between the two companies. MidnightTrader.com began

streaming its customized news feed on Market XT’s Web site in late

summer.

▼ New traders looking for pointers may want to check out

TradersCoach.com, an educational Web site focusing on aspects of trad-

ing from system development to psychology. Bennett McDowell, who

is a trader and manages many active trading accounts, developed the

site and has published “The Survival Guide for Traders,” another tool

for novice traders with tips on how to set up and organize a trading

business. For more information visit www.traderscoach.com.

▼ CyBerCorp has released its CyBerX2 platform, which combines fea-

tures from its CyberX product and CyberTrader including charting

capabilities and streaming Level II data. (CyBerX2 can be ordered

with the Level II data feed function deactivated.) With CyBerX2, users

also have the ability to trade options from a real-time streaming

option quote chain. A complete list of options for each individual

month is displayed, along with the price, symbol, net change, bid and

ask. Visit www.cybercorp.com for a demo of the product.

▼ Online broker RJT.com has added free real-time streaming quotes

to its client offerings. With “Jet Stream,” users can program up to nine

pages of quotes containing a total of more than 200 individual stocks

and choose whether quotes appear in fractions or decimals.

Fundamental data, such as 52-week, high-low price and date, P/E

ratio, EPS and ex-dividend dates, and a customizable toolbar are fea-

tures, as is an alarm that signals when pre-programmed levels are

reached. For more information go to www.rjt.com.

▼ BigEasy Investor (www.bigeasyinvestor.com) is a free screening,

charting and investment support program intended for both novice

and seasoned traders/investors. Among the product’s features are

two screening tools: Easy Screening, with which users can select from

more than 120 pre-defined screens using popular criteria; and Big

Screening, with more than 300 configured screens that users can

choose “as-is,” or customize. Another component is BigEasy

Investor’s Candlestick Guru, which recognizes more than 70 signifi-

cant candlestick patterns and paints a red box around the pattern for

easier identification.

▼ Paritech, a provider of tools for traders/investors in Asia, has

entered the U.S. with an office in San Francisco. The company is dis-

tributor for analysis programs such as Meta Stock and OmniTrader

and also runs software training courses globally. For more informa-

tion, visit www.paritech.com.

Send your new product information to: Amy Brader, Managing Editor or Jeff Ponczak, Associate EditorActive Trader Magazine555 West Madison, Suite 1210 • Chicago, IL 60661

Fax: (312) 775-5423

It’s more than a little ironic that thecommunications technology thre a t-ening the traditional brick-and-mortar exchanges is actually on the

v e rge of making electronic trading a more“human” experience. While pro p o n e n t sof open-outcry auction markets arg u eface-to-face is still the best way to do busi-ness, new developments in re a l - t i m eaudio, video and voice re c o g n i t i o n ,among other advancements, promise togive online trading a much more person-al — and sophisticated — touch in thenot-too-distant future .

The picture of today’s short-termtrader is primarily that of anindividual tethered to a com-puter screen during re g u l a rmarket hours, scrolling throughquote screens and charts, andpointing, clicking and peckingaway at a keyboard to enterorders.

But if you listen to PhilipB e r b e r, founder and chairman ofd i rect-access broker CyberCorp,a trading future filled withh a n d s - f ree technology,i n c reased mobility and gre a t e raccess to round-the-clock globaltrading is not far off. Incre a s e dintegration of brokerage servic-es, data and analytical tools —fueled by fatter bandwidth —will bring the markets mored i rectly to individuals. Tr a d e r swill effectively be wired into a

“virtual trading room” that gives themd i rect access to not only stocks, butoptions, futures, currency and intere s trate markets — around the clock, nearlyevery day of the week.

B e r b e r, who delivered the keynoteaddress on emerging trading technologyat the Online Trading Expo in Ontario,Calif., in August, portrayed a rapidlyevolving industry that will be driven bythe need to bring ever-more sophisticat-ed services to an increasingly demand-ing trading public. Among the develop-ments he sees on the horizon:

• Improved voice-recognition soft-

w a re technology (“natural languageparsing”) that will allow traders to enterorders by voice.

•A multi-stock Level II screen.•Amove away from tabular, text mar-

ket information toward more graphical,“cognitively intuitive” data displays.

• Real-time chat and audio that willallow you to do things such as commu-nicate with your bro k e r’s support inreal-time without a phone, as well asinteract with other traders.

• Expanded wireless technology thatwill give traders more mobility,enhanced by integration with the afore-

mentioned voice-re c o g n i t i o nt e c h n o l o g y. Going on theroad? Talk into your PersonalDigital Assistant (PDA) toenter your order.

• Integration of advancedscreening and analysis toolsthat will allow you to findstocks and other instrumentsthat fit your specific funda-mental and technical criteria,in real time.

• Enhanced conditionalorder capability, and intelli-gent order entry and routingfunctions will allow you toplace orders away from themarket based on variousprice, technical, or fundamen-tal factors, further freeing youfrom your computer screens.

Some of these develop-


T E C HF R O N T I E R ST E C HF R O N T I E R SWhat will trading be like a year from now?

Emerging technology promises to plug traders into the markets more directly than couldhave been imagined just a few short years ago. We’ll take a look at some of the new

developments on the hardware, software and communications front.

Neovision Heatmap

BY MARK ETZKORN AND THOMAS STRIDSMAN

Hardware • Software • Communications

ments, such as real-time chat supportand wireless technologies, are already inuse, in one form or another. Others, suchas a new generation of voice-recognitionsoftware that doesn’t require the exten-sive “training,” and which can processspeech re g a rdless of accent or othersmall idiosyncrasies, may be rolled outin the next 12 months or so.

The technology that made theE l e c t ronic Communication Networks(ECNs) and direct-access brokers possi-ble has had a profound impact on thetrading industry, Berber says. For onething, technology is blurring the linesbetween traditional exchanges andECNs (as evidenced by several ECNs’ongoing efforts to acquire exchange sta-

tus and Archipelago’s merger with thePacific Stock Exchange). The trend isirrevocably toward integrated, electronicmarkets.

In fact, Berber says, bringing near-total access to the global markets to theindividual trader’s desktop is not somuch a matter of solving technical prob-lems but of cutting through variousforms of red tape among diff e re n texchanges and governments.

“The technology hurdles are not assignificant as the regulatory and compli-ance ones,” he says. “Now you’re notjust dealing with the NYSE and Nasdaq,

but with exchanges and regulatory bod-ies around the globe.” As a result, Berber

thinks round-the-clock access tostock, option, currency and futuresmarkets from a single trading plat-form is likely still several yearsaway.

The latest software and hardwareon display at the Expo backed upB e r b e r ’s contentions. Obvioustrends include the continued pro-liferation of direct-access brokers,m o re widespread use of high-bandwidth online and networkconnections, integration of analyti-cal tools and trade execution, andnew ways to distribute and displaymarket data. We picked out a fewrepresentative products.

One item that typifies the spreadof simplified, intuitive graphicalrepresentation of market action isthe Heatmap by Neovision( w w w.neovision.com). The pro-gram uses different grids of color-

coded boxes that signify how much amarket is up or down, or the level of

v o l a t i l i t y, or some other measurableindicator (see p. 30). The theory is thatit’s easier and quicker to absorb informa-tion this way then to have to read thetypical text quote.

For example, you can set up a portfo-lio, and click to see percentage moves inthe diff e rent stocks. The darker blue a boxis, the more that stock is up; the darkerred a box is, the more the stock is down.Click on the stock and you get a detailedversion of the stock’s stats. Neovisiona l ready has versions of the software forB l o o m b e rg, Bridge/Telerate, eSignal andReuters, and it can run as a stand-alonep rogram as well.

Another color-coded price-changedisplay from CyberCorp’s CyberX2 trad-ing platform is shown below.

Stock search/screening functions arebecoming increasingly familiar addi-tions to direct-access broker analyticaltoolboxes and financial Web sites, butmany of them are rudimentary and mostof them (especially the Web versions)function on a delayed basis. You plug ina few criteria — volume, price level, per-haps a fundamental input such as P/Eratio or information on a technical indi-


FloorPass

MarketMatix

Tech round-up

Product Company Web address Phone

Heatmap Neovision www.neovision.com N/A

MarketMatix MarketMatix www.marketmatix.com (949) 608-0140

FloorPass Velocity Trade www.floorpass.com (877) 953-5274

PocketGenie WolfeTech www.wolfetech.com (909) 596-2700

HandTrade HandTrade.com www.handtrade.com (631) 592-1370

Visor Deluxe Handspring www.handspring.com (650) 230-5000

cator or two — and you get a list ofstocks that, based on yesterday’s closingprice, fit your profile.

MarketMatix (www. m a r k e t m a t i x . c o m )is a stand-alone screening tool thatsearches the entire stock market and,using a set of proprietary algorithms,finds stocks that are gaining momentumand streams the real-time results direct-ly to you through a pop-up window.You see the direction the stock is mov-ing, the intensity of the move, the timeof alert and the criteria being used togenerate it (see p. 31). Of greater interestto many traders is the company’s plan toallow users to set up customizablesearches in addition to the product’sbuilt-in screens.

MarketMatix is currently offering af ree 10-day trial. Subscription to theservice costs $48.88 per month (twomonths free when you pay upfront for ayear).

One company standing at the cross-roads of high-tech communications andthe traditional trading floor is FloorPass( w w w.floorpass.com), which caters toactive stock index futures traders.FloorPass offers a complete tradingsetup featuring dual flat-panel monitors,wireless keyboard and mouse, and digi-tal camera (see p. 31). The setup eschewsthe Internet in favor of a private network— you connect directly to them, bypass-ing potential ISP connection glitches. Inaddition to its more traditional quoteand charting capabilities (courtesy ofeSignal), the product features a live“squawk box” connection to the S&Pfutures pit and direct-to-the-pit orderentry (to the S&P, Nasdaq and Dowfutures).

The digital camera is used to enabletwo-way live audio connection withcompany support. A window opens up

so you can talk in real-time withFloorPass help when you have aproblem. The setup, technically, isfree. You get to use it when youopen an account with an affiliatedbroker.

I m p roved wireless technologywill make mobile trading more re l i-able than it’s been in the past. If youneed to take your trading on theroad, you might want to look atPocketGenie from Wo l f e Te c h( w w w.wolfetech.com). PocketGenie

is a software application that pro v i d e sw i reless Internet access to your text-based pager (see above, left).

So, in addition to staying in contactwith your near and dear ones and pro-viding up-to-the-minute news alerts,

your pager can connect to your onlinebroker for direct access to your account,w h e re you can get real-time quotes,track your portfolio holdings and evenplace orders. All communication usingthe PocketGenie is encrypted to ensuresecure transactions. Unlimited service is$14.95 per month.

Earlier this summer, WolfeTechalso announced the addition of aWAP- (Wi reless A p p l i c a t i o nProtocol) based browser fortwo-way pagers and otherPDAs, which essentiallywill give your hand-held device the sameb rowser capabilitiesas your computer.

H a n d Tr a d e . c o m( w w w. h a n d t r a d e . c o m ) i staking things one step furtherby providing the answer to a dayt r a d e r’s common question: “If Ineed to go to the bathroom, should Iexit all my trades or take my chances?”

With HandTrade’s PDA-based soft-ware, it doesn’t matter. The software will

allow you to trade in real time from vir-tually any location, turning your PDAinto a trading platform complete withs t reaming Level II, portfolio trackingwith alarm capabilities, charting capabil-ities and order placement with anyd i rect-access brokerage company. Tostay connected, you will need to hook upyour PDA to your cellular phone or to aw i reless modem. Curre n t l y, with theproduct in beta testing, the service ispriced at $50 per month. The future tar-geted market is an estimated 400,000professional and 8 million part-time daytraders around the world.

In the near future, HandTrade.compredicts that you should be able to per-form the same functions on your video-enabled, third-generation wire l e s s

phone with full multimedia capability.Despite the limits set by today’s cellu-lar phone and PDAstandards, the pro-gram seems very user-friendly with allthe necessary features in all the rightplaces.

Speaking of Palm-compatiblePDAs, check out the new Vi s o rDeluxe, from Handspring(www.handspring.com). It’s priced at$249, the same as the Palm IIIxe, butwith a few more features added

(including expandable memory and abuilt-in microphone), and slicker designand add-on hardware. Visor runs on thesame OS as the regular Palm Pilot andcan, therefore, make use of the same pro-grams provided by a wide variety ofthird-party program vendors. Check outw w w. p a l m g e a r.com, where you canshop for more than 220 finance-relatedproducts, ranging from Excel-compati-

ble spreadsheets to re a l -time stock trading,and thousands ofother pieces of soft-ware.Ý


PocketGenie

CyberX2

HandTrade

O ne of the interesting things about beinginvolved in the start of a new industry is thatyou get to witness its various growth stages.The last couple of years have seen a dramatic

increase in the use of direct-access trading systems, as well asefforts to educate traders on how to use these trading plat-forms.

Previously, this was done via books, seminars or onlinecourses. Now, however, the Online Trading Academy (OTA)has moved into the realm of personal multimedia instructionwith its “Fundamentals of Direct Access Trading” CD-ROMcourse. The training course is a scaled-down version of OTA’sone-week Boot Camp Level II trading course.

Any computer with a CD-ROM or DVD drive can run thecourse. The program is completely independent and requires

no downloading (unless you purchase itas a download from the Internet). It’sdivided into 12 training modules thatcover everything from market history,order routing, Level II screens, time andsales, and order types, to risk manage-ment, short selling, charting and tradingstrategies. Each section lasts about 40minutes and has a quiz at the end so youcan gauge your understanding of thematerial. OTA uses TradeScape’s direct-access trading interface throughout thecourse.

OTA makes good use of interactivemultimedia technology. The graphics arecrisp, and the audio commentary is clear.Each section that a student has to under-stand is given its own module. This sepa-ration makes it easier to learn the screensindividually and not be distracted byother information they will learn later.While the speed on the trading screens isgreatly reduced, it nonetheless gives agood indication of how real trades lookand feel on a direct-access screen.

The short quizzes in each sectionrequire you to answer before proceeding


Hardware • Software • Communications

BY KIARA ASHANTI

Online Trading Academy’s direct-access trading multimedia course spans 12modules and topics ranging from order entry to risk management.

Product: Fundamentals of Direct Access Trading, version 1.0

Company: Online Trading Academy (www.tradingacademy.com)

Phone: (888) 841-8418 or (949) 930-2088, 6:30 a.m. to 7:00 p.m. (PST)

Online contact/customer service: [email protected]

Price: $295

Required system: Windows: Pentium 133, 32 MB RAM16-bit color display Windows 95/98/NT/2000 or higher.Screen display should be set to 800x600 or higher.Macintosh: Power PC, G3 CPU RUNNING MAC OS 7.0 orhigher, 32 MB RAM. The program runs off the CD, sohard drive space is not necessary unless program ispurchased via the Internet as a download.

Product R E V I E W:

Fundamentals of Direct Access Tr a d i n g

to the next screen page. This not only ensures that you areabsorbing the information, but also keeps you from feeling asif you’re being “talked at” by your computer. If you don’tunderstand something or missed a question, you can click on a“back” button to repeat the previous information.

Sprinkled throughout the CD-ROM are demos showing 30-second clips of OTA trainers putting through orders in real-time. However, the movement and commentary might be toofast to follow for novices or traders not accustomed to direct-access screens.

Additional features include a help menu that lists majormarket makers, a glossary of financial terms and “trader”slang, a very helpful list of economic indicators, a trading rulerthat converts fractions to decimals and a template for trackingyour trades. (While the glossary is extensive, it contains wordsthat are not discussed in the modules, leaving you no contextas to how it pertains to direct-access trading.)

One notable section of the help menu that could beimproved is a list of common trading errors. The list is in writ-ten form; this section would have had more impact as a sepa-rate module with audio examples. Finally, while the informa-tion in a video portion that provides quick advice on varioussubjects from OTA trainers and traders is valuable, the video isa little choppy and skips a beat like a scratched record.

Ultimately the only thing that matters is whether this CD-ROM delivers its promise as a good training tool. And theanswer to that is a strong “Yes.” OTA has placed enough toolsand information on this product to make it a very useful guidefor novice traders. The explanations on how to use the tools ofa direct-access trader are clear and understandable. The graph-ic examples of what to look for when trading are great (withthe exception of the demos). In keepingwith OTA’s overall focus on risk manage-ment, the moderator continually showshow a trade can turn bad, and the finan-cial consequences that can result if it does.Another strong point is the program’sthorough explanation of margin — whatit is, how to use it and how it can hurtyou.

“Fundamentals of Direct A c c e s sTrading” won’t turn you into a star trader,but that’s not its purpose. It gives begin-

ners enough information to know what to look for as theybegin paper trading on a direct-access system.

So if you’re looking to get into day trading, or you are think-ing about switching from a regular online broker, this productis a good beginning supplement to the reading and seminarsyou’ll need to make it in the direct-access trading world.Ý


O TA has placedenough tools andi n f o r m a t i o n on thisproduct to make ita very useful guide fornovice traders.

The course uses TradeScape’s direct-access interface to illustrate various trading concepts.

PRODUCT SUMMARY

Product: Fundamentals of Direct Access Trading, version 1.0

What it is: Multimedia training course that is asimplified version of the Online Trading Academy’s(OTA) Boot Camp Level II trading course. It is dividedinto 12 modules, each approximately 40 minutes long,totaling eight hours of instruction.

Who the program is for: Beginners

Upside: Simple graphics and straightforward audioinstruction offer beginners excellent commentary onhow to read Level II screens, time and sales, andorder entry. Humorous animation keeps the materialentertaining, as well as offering good analogies ofkey concepts and trading dangers. Lessons aregrouped together logically, with concise, clear instruc-tion. Extensive economic and trading backgroundinformation.

Downside: Very few “real-time” examples of trainingconcepts (i.e., watching prints and Level II screenmovement). Demos provided might be too quick for abeginner to follow. Video clips are a little choppy.Traders with slower computers beware.

O ver the past few years, thousands of potentialtraders have set up shop at day-trading firms— only to have their capital wiped out in a mat-ter of weeks or months.

It would almost seem that you could simply study what thesetraders did and do the reverse. Think of it this way: How manytimes have you said, “If I had just done the opposite of what Iactually did, I would have made money?”

If you are an avid Level II user, this may be a valid question.Trading is difficult no matter how you look at it, but if you arelearning to do the wrong things, it can make it next to impossi-ble to make money.

What you will soon discover is that some of the “basics” cur-rently being taught about Level II are wrong. We’ll take a look atsome of the realities of using Level II quotes and outline a fewtrading strategies that are rarely discussed in any books or trad-ing classes. In fact, just one of the ideas that will be revealed willincrease the odds of a winning trade dramatically. But let’s startout reviewing some of the basics.

The basic premise of Level II is very simple. Figure 1 shows anexample of a Level II screen. The left side of the screen containsinformation about who is attempting to buy stock. The right sideof the screen shows who is attempting to sell. On both sides ofthe screen you can see at which price each participant is buying


Make no mistake, Level II is the market

makers’ turf. If you want to play on this

field, you need to arm yourself with the

proper knowledge. Here’s a look at how

you can stay on the side of the “smart

money” and avoid becoming a victim of

the market makers’ games.

What you see is (not) what y ou ge t

BY DEWEY E. BURCHETT

TRADING Strategies

or selling, and what size (how manyshares).

In addition to market makers such asMerrill Lynch (MLCO), First Boston(FBCO) and Goldman Sachs (GSCO),E l e c t ronic Communication Networks(ECNs) such as Island (ISLAND/ISLD),A rchipelago (ARCA/ARCHIP) andRedibook (REDI) are re p resented on theLevel II screen. These ECNs are primarilyused by day traders and active traders.Keep in mind that a market maker is alwayson both sides of the market, although obvi-ously at diff e rent price levels.

With the basic premise of Level II inmind, it would be easy to jump to theconclusion that if you knew what theorder flow was going to be over the nextfew minutes based on what was on thescreen, making money would be likeshooting fish in a barrel. For example,you might assume that if seven marketmakers were stacked up at high bid, say50, with only one market maker on theoffer at 50 1⁄16, the stock is going up

because of the order flow. Conversely, ifyou saw market makers stacking up onthe offer or sell side, it would only makesense to position yourself for a short, orsell a stock that you were long.

For the new trainees at the local daytrading firm, this makes total sense, andthey can hardly wait to get theiraccounts up and running. If the Level IIs c reen could really be taken at facevalue, then theoretically it would take agreat deal of effort not to make moneygiven the available information.

Unfortunately, things are not alwaysas they seem.

Traders need to understand one conceptwhen using Level II: Market makers letyou see what they want you to see. Inaddition, always remember that everyonesees the same thing that you are seeing.

Market makers have two main objec-tives: Execute order flow at the best pricefor their big clients, and make money

scalping the stock for the firm. You mayhave heard that the “smart money” betsin the last few minutes before a horserace. If someone knows his horse isgoing to win, he is not going to bet bigearly and tip off all the other handicap-pers. If he did, other players would betthe same horse, which would lower theodds and the payoff.

Just as the astute handicapper doesnot show his hand, neither does the mar-ket maker. In fact, it is in the best interestof the market maker to make a strongstock look weak on Level II, and viceversa.

What often ends up happening is thata stock with buying interest looks weak,and a stock with selling interest looksstrong. In other words, on Level II youwill see market makers stacked uptogether at the top of the offer, with veryfew bids — right before the stock makesthe next move to the upside. Again, thereason it looks this way is because themarket makers want to make the market

look weak when theyhave stock to buy so theycan get a good price. Theywould prefer to make thestock look weak so traderswill either sell to them onthe bid or lower the cur-rent offering price, inwhich case they will buyit from the new seller.Often they will pull anoffer when some of thestock is taken out at thatprice level.

As soon as you under-stand that the true inten-tions of the market mak-ers usually are not reflect-ed on Level II, you can usethis information, in con-junction with other trad-ing strategies, to deter-mine the proper entry andexit points for the stocksyou are trading. Forexample, assume youwant to go long a particu-lar stock during the day.You decide that youwould like to buy on thenext pullback based on atechnical indicator youwere using.

If you had listened tomuch of the popular


Source: Real Tick III, Townsend Analytics

In the pre-market, two market makers, MASH and MWSE, are on the inside offer, but thebid side of the market is stacked up with ECN/day trader orders.

FIGURE 1 LEVEL II STANDOFF

advice regarding Level II, the last thingyou would do is buy a stock when mar-ket makers were stacked up on the bestoffering price. Under such circumstancesyou would probably think there is toomuch selling pressure. But this is theincorrect approach. Instead, you shouldadhere to your trading signal and makethe trade. A good way to play the longside, keeping in mind the tendency ofmarket makers to not reveal their trueagenda, is to try to take advantage of thismarket maker “bluff” by putting in a bid(instead of just lifting the offer). Manytimes more inexperienced traders willhit your bid when they see the marketmakers stack up on the offer (this worksbest using an ECN like Island). This willgive you a better entry price, as you areusing the market makers’ trading tacticsto your own advantage.

Keep in mind, however, there are cer-tainly times, such as during heavy trend-ing, when a preponderance of marketmakers on either side of the marketreflects true buying or selling interest.Learning how to read market makeraction is something that is acquired overtime. The message here is to rememberthat you can’t always take what you areseeing on the Level II screen at facevalue. The more liquid a stock, the morelikely you will see a better indication of

the true buying or selling interest.The opposite approach would be used

for trading on the short side, althoughentry is more difficult because of theuptick rule. Viewing Level II from am o re realistic standpoint — re a l i z i n gthat what is on your screen is exactlywhat market makers “want” you to see— will help you minimize losses andenable you to not get thrown off whenusing other trading strategies. It isimportant to keep in mind, however,that this information should not berelied on exclusively; it should be usedin conjunction with other eff e c t i v estrategies. Also, illiquid stocks are lessreliable as there are fewer market partic-ipants, and one “big gun” can ultimatelycontrol the stock (see “Cutting in front ofthe ax,” Active Trader, September).

T h e re is one particular strategy thattends to yield high-probability tradesetups, although these opportunities canbe difficult to find at times. Theapproach is based on the premise thatnot only do the bigger market makersultimately control which direction astock will go, but that day traders, whenthey act collectively, get on the wrongside of the market at the same time. Inthis context, day traders represent the“dumb money,” to use trading terminol-ogy. There is an easy explanation for thisphenomenon, as the following examplewill illustrate.

Dumb money usually comes togetheraround news events such as new prod-uct developments, strategic allianceswith other companies and the generalfancy press releases that don’t re a l l yamount to a hill of beans. In the excite-ment surrounding the news, and com-pelled by the greed of making a quickbuck, traders will chase the latest hotstock of the day using the ECNs. Whenthis happens, most of the market makerswill raise their offers and make thetraders pay up for the stock.

Once the chase is on, and day tradersbegin to outbid each other, the stock maymove up swiftly, although it generallywill do so on light volume. What endsup appearing on the Level II screen ismostly ECNs on the bid side, accompa-nied by very few market makers as buy-ers. Most of the time you will see sometrades go through on the offer as well,but few of any significant size, and thedown move that follows can be evenmore violent than the preceding rise asday traders bail out on the stock.

To find these potential trades, keep trackof stocks with news prior to the market


M a ny times more i n e x p e rienced traders will

hit your bid when they see the market makers stack up

on the offe r. . .

General Magic (GMGC), daily

Previous trading defined resistance in GMGC, and a likely place to look for ashort trade opportunity

FIGURE 2 RESISTANCE DEFINED

Source: Qcharts by Quote.com

12 19 26 3 10 17 24 31 7 14

9 1⁄2

9

8 1⁄2

8

7 1⁄2

7

6 1⁄2

6

5 1⁄2

5

4 1⁄2

4

5 7⁄8

July Aug.

open. These stocks often will trade upsignificantly in the pre-opening market,and have short-lived runs in the first fiveminutes of trading as market makers sellthe stock to an investing public that onlybuys when the market opens. Othertimes these stocks will fall immediatelywhen the market opens.

The best way to determine how toplay these situations is to watch the vol-ume. If the volume continues to beheavy all the way to the opening bell thestock will likely continue to push slight-ly higher within the first five to 10 min-utes of trading. If the volume is relative-ly light, the stock will most likely imme-diately get hit with pro f i t - t a k i n g .Fortunately, you do not have to wait foran uptick in the pre-market to sell the

stock short; waiting until after the openof trading will require you to sell on anuptick. Although other news-re l a t e devents occur throughout the day, mostsignificant news is released before theopen, so you have to do your homeworkearly in the day.

F i g u re 1 shows a Level II screen ofGeneral Magic, Inc. (GMGC). On thisparticular day, GMGC signed an agre e-ment with IBM. And as usual, daytraders were hard at work bidding thestock up in the pre-market, as evidencedby the large number of ECN (ISLD andARCA/ARCHIP) bids on the left handside. Also notice that there is only onemarket maker (SHWD) on the bid side,two levels down, which means they arenot aggressively buying the stock. In

addition, most market makers are on theo ffer side of the market (right side of thes c re e n ) .

N o w, who will ultimately win thiss t a n d o ff? This example shows the time of9:19 EST — 11 minutes prior to the mar-ket open. The first question when spot-ting a situation like this is to figure outw h e re the resistance levels are. Figure 2shows that General Magic has re s i s t a n c e(prior support) in the 6 to 6 1⁄2 a rea . Itwould be logical to expect the stock tomeet resistance in this area once itopened; this is where you would look tosell the stock short, given that re l a t i v e l yheavy volume continued to the openingbell.

In this case, volume continued tocome in, and within 30 seconds the stockgapped open to the upside and made thehigh of the day at 6 7⁄16. Although above-average volume continued in the first 15minutes, the intraday chart in Figure 3shows the stock quickly pulled backfrom the high. From then on it traded aslow as 5 3⁄4, but managed to close at 5 7⁄8 onthe day. For traders looking to scalp aquarter-point or more, or 4 to 6 percent,this was a quick, low-risk play.

One other misconception traders haveis that when they see a big buy or sellorder, say 10,000-plus shares, they auto-matically assume it dictates buying orselling pressure. An example would bethat you see a buyer (generally a marketmaker or Instinet) come in with an orderof 10,000 shares on the bid side. Traderstoo often assume this means the stockwill immediately rise as the buyerattempts to buy shares. This is generallyan incorrect assumption, and it all goesback to the fact market makers are let-ting you see what they want you to see.

If you wanted to buy 10,000 shares,would you advertise your intentions tothe world so they would make you payup for it? Of course not, and in most sit-uations neither will anyone else. If youreally look at such bids and offers, theyare generally a couple of levels downfrom the top of the bid or offer, and usu-ally only appear on the screen for a fewseconds. In other words, someone has anagenda, and it’s not necessarily one withyour best interests in mind. These trad-ing scenarios should be viewed as warn-ing signs. If you understand that thingsare not as they seem, you can better prof-it from what really is.Ý


The stock quickly established the high of the day (at 6 7⁄1 6) and began to decline.

FIGURE 3 GAP OPENING

General Magic (GMGC), 5-minute

6 1⁄2

6

5 1⁄2

5

4 1⁄2

0

Aug. 15,2000 Aug. 16,2000

10:00 11:00 12:00 13:00 14:00 15:00

10000000

In the excitement

of the new s,and compelled

by the greed of making

a quick buck,traders will

chase the latest hot stock

of the day using the ECNs.

T rading hot IPOs on the firstday they go public can beboth very intimidating andvery profitable. If you use

an extra bit of caution, they can lead toextraordinary gains in extremely shorttime periods. The strategy consists ofpinpointing a specific pattern: enteringIPOs on an intraday pullback of the ini-tial thrust that often occurs when theybegin trading.

Use 10-minute charts to identify tradesetups. First, though, keep in mind thatnot all IPOs are worth trading; youshould only focus on those issues thatimmediately rally from the instant theyopen for trading. To enter, look for a pull-back consisting of at least two bars withlower highs than the preceding bar thatdo not break the low of the trading ses-sion. The lower highs do not have to beconsecutive and there does not have to bea minimum number of bars that should

pass before the two lower highs occur.Buy one tick above the high of the pre v i-ous bar. An initial stop-loss is placed onetick below the low of the pullback.

Because these new issues are so volatile,you should risk a little less than you nor-mally would on a trade. For example, ifthe high on which you are basing yourentry point is 44 (44 1⁄16 entry) and the lowof the pullback is 43 (42 15⁄16 stop-lossexit), then you would be risking 1 1⁄8points on the trade and would adjust thenumber of shares accordingly, based onwhat dollar amount you wanted to riskon the trade.

Because you will probably encounterexcessive slippage on your entry and exitin these trades, you should only tradet w o - t h i rds of the shares you would nor-mally trade. For example, if you are risk-ing $2,000 on this trade you would onlybuy (rounding down to the nearest 100s h a res) 1,100 shares ($2,000/1.125 *.66).

A word of caution: Because speed of exe-cution is critical, it’s advisable to tradethis strategy only using a direct accessbroker. Hot IPO stocks can make huge


Hot IPOs must be traded with caution, but they

can bring substantial gains for nimble traders.

Intraday pullbacks can get you in

on big moves on the first day of trading.

Opening day O P P O RT U N I T I E S

TRADING Strategies

BY STEVE WENDLANDT

T

Opening day

Because hot IPOs are so volatile, you should risk a little less on t h e s e p o s i t i o n s t h a n you normally would on a t r a d e .

moves in short periods of time; tradingsuch issues can be extremely riskyunless your executions are almostinstantaneous. The delays in executionsand confirmations typical of standardWeb-based online brokers are toolengthy for this kind of trading.

One of the simplest and best moneymanagement strategies you can use isthe “2-for-1” approach. (Giving cre d i tw h e re credit is due, Larry Connorscoined this style of position managementin Connors on Advanced Tr a d i n g .)

It works like this: When you have aprofit equal to your initial risk, sell halfyour position and move your stop onthe remainder of the position to breakeven. In the previous scenario, the initialrisk was 1 1⁄8 points, so when the stockadvanced 1 1⁄8 to 45 3⁄16, you would sellhalf the position and move your stop tobreak even.

Then look for either a “parabolic”move to sell the remainder, or exit on theclose if you are not stopped out atbreakeven on the balance. A parabolicmove in this case would take the form ofan extremely large-range bar accompa-nied by high volume.

Let’s look at some examples. On July 25,Blue Martini Software (BLUE) openedfor trading at 431 ⁄2 and immediatelybegan to rally (see Figure 1). Bars six andseven were back-to-back bars with lowerhighs that did not exceed the low of thesession. Along position was triggered at44 1⁄4 on bar 8, one tick above the high ofthe previous bar. The initial stop-losswas at 43 7⁄1 6, one tick below the swinglow of 43 1⁄2, giving the trade an initialrisk of 13⁄1 6. Using the 2-for-1 appro a c h ,when the stock reached 45 1⁄1 6 (13⁄1 6 a b o v ethe entry price) half the position wassold and the stop moved up tob reakeven on the re m a i n d e r.

At this point you would look for anexpansion (wide-range) bar, which canbe somewhat subjective. Bar 16 was anexcellent expansion bar (large rangewith high volume) and would have pro-vided a good exit point for the balanceof the position.

F i g u re 2 illustrates another example.On July 13, Sunrise Telecom (SRT I )opened for trading at 31 and immediate-ly rallied, trading as high as 40 in the first

58

56

54

52

50

46

44

42

60,000,000

40,000,000

20,000,000

541,000

57 5⁄32

Volume

11:40 12:10 12:40 13:10 13:40 14:10 14:40 15:10 15:407/25 Tuesday

Source: QCharts by Quote.com

This IPO strategy should only be used on stocks that immediately rally whentrading begins. Entry occurs after the stock makes two lower highs that donot drop below the low of the session.

FIGURE 1 INTRADAY IPO PULLBACK

Blue Martini Software (BLUE), 10-minute

Enter one tick

above previous high

Initial stop one tickbelow swing low

Expansion bar

1

16

2

3 4 5 67 8

48

40

36

32

2,000,000

1,000,000

188,400

44 1⁄4

Volume

12:10 12:40 13:10 13:40 14:10 14:40 15:10 15:40 09:40 10:007/14 Friday


Controlling risk is essential. The initial stop is placed at the swing low of thepullback. When the trade has a profit equal to the initial risk amount, halfthe position is sold and the stop on the remainder moved to breakeven. Thebalance of the position is sold on an expansion (wide-range) bar, or held untilthe close.

FIGURE 2 MANAGING THE POSITION

Sunrise Telecom Incorporated (SRTI), 10-minute

Enter

Sell half the position and movestop on remainder to breakeven

Initial stop

Expansion bar

1

11

5432


20 minutes of the session. Bars 3 and 4both made lower highs than the pre c e d-ing bars, but the highs did not bre a kbelow the low of the session.

A long position was entered on bar 5at 37 13⁄1 6, one tick above bar 4’s high of37 3⁄4. The initial stop was placed at 36 3⁄1 6,one tick below the swing low of thepullback (the low of bar 5) at 36 1⁄4, for arisk of 1 5⁄8 on the trade. Using 2-for- 1money management, half the positionwould be sold at 39 7⁄1 6 (1 5⁄8 above theentry price) and the stop on the re m a i n-der of the position moved up to theb reakeven point.

Bar 11 presented an opportunity toexit the balance of the position on anexpansion bar with increased volume(bar 10 was also a possibility, althoughvolume was not as high on this bar).Alternately, holding the trade for the restof the session would have resulted in anexit at the closing price of 40 1⁄4.

On a week when there are many IPOs,

you should see several setups. However,please use caution, because there is noother time when a stock is more danger-ous to trade than the first day an IPOgoes public. Proper money managementis paramount.

A third example occurred in DyaxCorp. (DYAX) on Aug. 15. After openingat 19 and rallying to 22 1⁄4 in the first bar,the stock made three consecutive lowerhighs on bars 2, 3 and 4 that did not dropbelow the session low. A long trade wastriggered on bar 5 at 21, one tick abovethe bar 4 high of 20 15⁄16. The stop-loss wasplaced at 19 3⁄8, one tick below the bar 3swing low of 19 7⁄16 , giving the trade aninitial risk of 1 5⁄8. Using 2-for-1 moneymanagement, half the position would besold at 22 5⁄8 (1 5⁄8 above the entry price)and the stop on the remainder of theposition moved up to the bre a k e v e npoint. Because a definitive expansionbar did not appear, the position was exit-ed on the close.Ý

Hot IPO stocks c a n make huge moves i n short periods of time; trading such issues can be extremely risky

unless your executions are almost i n s t a n t a n e o u s .

Strategy snapshot

1. Only use the strategy on hot IPOs that immediately rally when trading begins.

2. Look for a pullbackconsisting of at least two lower highs on the 10-minute chart that do not break the low of the trading session.

3. Go long on a move one tick above the high of the previous bar.

4. Place initial stop-loss one tick below the low of the pullback.

5. Exit half the position when the profit is equal to the initial risk and move the stop on the remainder of the position to breakeven.

6. Exit balance of position on a parabolic move or on the close, whichever comes first.

30

28

26

24

22

20

2,000,000

1,000,000

280,200

25 11⁄16

Volume

10:40 11:10 11:40 12:10 13:10 13:40 14:10 14:40 15:10 15:40 9:40 10:10 8/15 Tues. 8/16 Wed.


Once half the position is exited, if no parabolic move occurs, hold theremainder until close.

FIGURE 3 HOLDING UNTIL THE CLOSE

Dynax Corporation (DYAX), 10-minute

Take profit on half of position,raise stop to break even

Initial stop

Exit on close

Enter

1

2 3

4 5


W hat makes a goodtrade? Well, in retro-spect, most traderswould say a nice prof-

it makes a good trade. But when you’re

putting a position on, the outcome isunpredictable. We’d all like to know atrade will be good in advance, but alas,the markets are not so accommodating.

What you look for when you’re get-ting in a trade is an entry point wherethe odds of a move in your favor are bet-ter than average. Then, by having a planthat determines when and where you’llexit with either a loss or a profit, you tryto structure a trade where the potentialreward is greater than the known risk.

The advantage of trading breakouts ofcongestion patterns such as tradingranges, triangles, flags and pennants isthat these formations allow you to clear-ly define the risk on your trades. Forexample, if a stock moves into a tradingrange after a rally, you may look to buyan upside breakout of the range in antic-ipation of a continuation of the uptrend.The logical place to put an initial protec-tive stop is below the low of the tradingrange, because a downside re v e r s a lthrough the support of the range wouldbe a bearish development.

F i g u re 1 provides an example. In lateJune, Microsoft (MSFT) established a re l-atively narrow trading range aftera p p roximately a 16-point rally. The stockb roke out of the upside of the range( a round 80 1⁄8) on July 6. The initial pro-tective stop would have been placed just


FIGURE 1 FALSE BREAKOUT

Microsoft Corporation (MSFT), daily

12 19 26 3 10 17 24 31 7July Aug.

Upsidebreakout

Stopped outSupport level used as initial stop

A trading range develops in the aftermath of a sharp rally. After an initialupside breakout, the stock reverses to the downside, stopping out the longposition.

82

80

78

76

74

72

70

68


72 5⁄8

More bang for your buck: PATTERNS

WITHIN PAT T E R N SHow to create trade opportunities with increased

reward and decreased risk by trading patterns within patterns.

TRADING Strategies

below the support level of the tradingrange, around 76 1⁄2. A move back belowthis level would suggest the upsidet h rust was actually a false breakout andthat the trade should be exited.

That’s exactly what happened. Twodays after entry the stock had pulledback into the trading range. It movedsideways to lower over the next severaldays before, on July 19, penetrating the

downside of the range and stopping outthe long trade.

The risk on this trade was a moderate3 5⁄8 points. But what do you do when atrading range is much wider and a stopbased on either the support or re s i s t a n c elevel re p resents too large a risk? Figure 2shows a much more volatile trading rangethan that in Figure 1. Using the samea p p roach as in the previous example —buying on an upside breakout of the trad-ing range and placing an initial pro t e c t i v estop below the low of the range — wouldre p resent considerable risk.

As a result, some traders place the ini-tial stop in the middle of the tradingrange. This more conservative method isbased on the idea that a strong breakoutmove should follow through immediate-ly and not reverse back into the tradingrange. Another way to reduce risk onbreakout trades is to look for shorter-term patterns within larger patterns thatallow you to place your initial stop-losscloser to your entry point.

When the risk implied by a particulartrading range is exceptionally large, youcan look for smaller congestion patternsnear the support or resistance levels ofthe range. Basing entry and stop pointson the levels defined by the smaller pat-tern can reduce the risk on the trade aswell as provide the opportunity for earlyentry into the position.

Figure 3 shows the formation of awide trading range in Oracle (ORCL) atthe beginning of this year.A trader look-ing to enter long on an upside breakoutof this range would have to accept a riskof more than 16 points, assuming thebottom of the range was used for the ini-tial stop-loss.

However, a much narrower tradingrange developed in February. Using thisrange as the basis of an upside breakouttrade would have offered the same entrypoint but a much closer stop. In thiscase, placing a stop one tick below thelow of the narrower trading rangewould have reduced the risk to 6 3⁄4points. For a short-term trader, this rep-resents a large stop, but it’s still a dra-matic improvement and the pro f i tpotential for the move out of the largertrading range is still intact. (Later, we’lllook at the practical risk-reward impactthis can have on a trade.)


FIGURE 2 RANGE RISK

International Business Machine Corp. (IBM), daily

4 11 18 25 1 8 15 29 6 13 27 3 10 24 31 7 14 28 6 13 20 27 3 10 24 1 8 15 22 30 5 12 19 26 3 10 17 24 31 7 14N o v. Dec. Jan. 2000 Feb. Mar. Apr. May June July Aug.

Using the opposite side of a trading range as a stop for a breakout trade canresult in large initial risk if the trading range is wide.

130

125

120

115

110

105

100

95

90


120 15⁄16

FIGURE 3 CONGESTION WITHIN CONGESTION

Oracle Corporation (ORCL), daily

Narrow range

Wider trading range

4 11 18 25 1 8 15 22 29 6 13 20 27 3 10 18 24 31 7 14 22 29 6 13 20 27 3 10 17 24 1 8 15Oct. Nov. Dec. Jan. 2000 Feb. Mar. Apr. May

A shorter, narrower trading range forms just at the resistance level of a larg -er range. Using the support level of the smaller range as a protective levelfor an upside breakout substantially reduces the trade’s initial risk.

90

80

70

60

50

40

30


77

{

Figure 4 provides another example.In this case, EMC Corp. (EMC) repeat-edly pulled back from resistance around72 1⁄2. Because a well-defined horizontaltrading range did not develop (the stockswung back and forth in an increasinglywider range), the most recent swing lowaround 51 would be the reference pointfor the initial stop-loss — a risk of morethan 20 points.

H o w e v e r, as the stock bounced offthat low and made another run at theresistance level, it formed a flag consoli-dation from June 7 to June 12 with a higha round 69 7⁄8 (the highs of the bars in theflags were within 1⁄16 of each other) and alow around 66 13⁄1 6. The upside bre a k o u tof this flag provided an early entry to thesubsequent surge that pushed the stockpast the 72 1⁄2 resistance level to newhighs.

F i g u re 5 shows a 15-minute chart ofthe Nasdaq 100 tracking stock (QQQ).The stock formed a large bottoming pat-tern (a head-and-shoulders bottom pat-tern; the preceding sell-off is not shown)with resistance around 82 5⁄8. As the stocka p p roached the resistance level for thesecond time, on May 30, it consolidatedin a narrow flag pattern with re s i s t a n c ea round 82 7⁄3 2 and support around 81 5⁄8.Playing an upside breakout of this pat-tern and using its support level for the


FIGURE 4 FLAG NEAR RESISTANCE

EMC Corporation (EMC), daily

Flag

Resistance

27 4 11 18 25 1 8 15 29 6 13 27 3 10 24 31 7 14 28 6 13 20 27 3 10 24 1 8 15 22 30 5 12 19 26 3 10 17 24 31 7Oct. Nov. Dec. Jan. 2000 Feb. Mar. Apr. May June July Aug.

A small flag forms just below a well-defined resistance level, offering earlyentry into the upside thrust move.

90 1⁄2

80

70

60

50

40

30


FIGURE 5 NARROW FLAG

Nasdaq 100 Index (QQQ), 15-minute

Narrowflag

S

H

S

Resistance

19 22 23 24 25 26 30 31 1 2May June

A narrow flag consolidation forms near the resistance level of an intradayhead-and-shoulders bottom pattern. The low of the flag provides a lower-riskstop level than the most recent swing low.

92

88

84

80

76

93 3⁄8


The advantage of trading breakoutsof congestion p a t t e r n s such astrading ranges, triangles, flags and pennants is thatthese f o r m a t i o n sallow you to clearlydefine the risk o nyour trades.

initial stop (rather than the most re c e n tswing low around 76) reduced the risk ona long trade to less than a point.

A final example is shown in Figure 6.H e re, in the middle of a larger tradingrange with resistance around 32 3⁄8,M o t o rola (MOT) formed a flag consolida-tion in late-October 1999 that off e red the

opportunity to trade an upside move withlower risk. The stock gapped out of theflag (a bullish sign) above 31 1⁄2 and contin-ued to run past the resistance of the larg e rtrading range. Placing a stop just belowthe flag support at 29 15⁄1 6 would havereduced the initial risk on the trade to lessthan two points. As was the case withF i g u re 4, the smaller pattern allowed youto both use a tighter stop and get in earli-er on an upside bre a k o u t .

F i g u re 3 provides a good example of howthis approach can work in the context of acomplete trade plan. The rally from thelate-October 1999 low to the early-January 2000 high was 41 7⁄3 2. The stockthen moved sideways, forming the larg e rtrading range. Atrader looking to buy on

an upside breakout of the range could usethe m e a s u red move a p p roach, whereby thesize of the previous price move is addedto the current price, to project a price tar-get. Adding the size of the price movep receding the trading range to the low ofthe larger trading range (around 46 5⁄8)results in an upside target of 87 27⁄3 2.

Using the measured move appro a c hon the smaller price swing from Jan. 28low of 465⁄8 to the Feb. 14 high of 64 3⁄4 ( 18 1⁄8points) sets up a shorter-term price targ e tof 77 7⁄1 6. This level would mark a good

spot to take at least partial profits on theposition and raise the stop on the balanceof the position. The stock actually formedanother flag after hitting a high of 76 1⁄2 o nFeb. 28. This consolidation marked anopportunity to exit part of the positionwith a profit; the stop on the remainder ofthe position could then be moved up tothe breakeven point, locking in a profit onthe trade. (For more information on tak-ing profits and moving stops, see“Opening day opportunities,” p. 42.) Thebottom line: The development of thesmaller trading range allowed the estab-lishment of a trade with a price targ e tbased on the larg e r, longer-term pricepattern with a risk based on the smaller,s h o r t e r-term price pattern.

Another general advantage of thisapproach is that it increases your flexi-bility. Even if you are stopped out on amove through the support of the smallercongestion pattern, you can still re-entera long position if the market reversesagain and breaks out above resistance asecond time. For example, a trader whowent long on the intraday upside thrustabove resistance (say, at 62 5⁄8) on Feb. 14and used the low of the smaller tradingrange (around 58 5⁄8) as the stop level,would have been stopped out on theintraday downside thrust on Feb. 22.However, as mentioned earlier, this lossis much smaller than the one that wouldhave occurred had the stop been placed

below the low of the larger tradingrange, which was nearly 12 points lower.

These patterns may develop relativelyinfrequently, but they fulfill the primarygoals of smart trading: They allow youto establish trades with shorter-term riskand longer-term profit potential. Inf u t u re articles we’ll expand on theseideas by looking at additional measuringobjectives and ways to put breakoutsinto context in relation to underlyingtrends of different magnitudes.Ý


FIGURE 6 EARLY ENTRY

Motorola, Inc. (MOT), daily

Flag

Trading range

3 10 17 24 1 7 14 21 28 6 12 19 26 2 9 16 23 30 7 13 20 27 4 11 18 26 1 8 15 22 29 6 13 20 27 3May June July Aug. Sept. Oct. Nov. Dec. Jan. 2000

A flag forms in the middle of a larger trading range. Even though pricegapped above the flag, playing the upside of this smaller pattern offeredearly entry and a tighter stop on a long-side trade.

52

48

44

40

36

32

28

49 171⁄256


When the risk implied by a particular trading range is exceptionallylarge, you can look for smaller congestion patterns near the s u p p o r tor resistance levels of the range.


J apanese candlestick chartshave become immenselypopular over the past 15years and are now a familiar

sight to most traders. However, consis-tent application and interpretation ofcandlestick patterns — especially whenused in conjunction with Western techni-cal analysis methods — have eludedmany traders. In this article, we’llexplain how to use Japanese candlestickpatterns with retracements and tests.

First, let’s define the concept of trend.An uptrend is a series of higher lows andhigher highs; a downtrend is a series oflower highs and lower lows. The higherlows of an uptrend are connected toform uptrend lines. The lower highs of adowntrend are connected to form down-trend lines.

A retracement can be defined as apullback in an uptrend or a bounce in ad o w n t rend. These retracements thengive rise to so-called tests: If the presentuptrend line is broken on a retracement(pullback) but buyers enter the market tomove it upward once again, this sets upa test of a top. If the present short-termdowntrend line is broken on a retrace-ment (bounce) but sellers show up tomove it downward once again, this setsup a test of a bottom.

Common Japanese candlestick pat-terns can be divided into two major cat-egories: reversal patterns, which signalthe beginning of a retracement in atrend; and continuation patterns, whichare found at the end of a retracementand signal the resumption of the trend.In our trading, we look for reversal pat-

terns on both a test of a top or bottom,and on a retracement (pullback orbounce).

Let’s look at a recent example of tests inthe popular Nasdaq 100 index-tradingstock (QQQ).

J A PANESE CANDLESTICKSTrading tests and retracements with

Candlestick patterns help highlight the short-term battle between the bulls

and the bears. Western trend analysis helps identify test

and retracement levels. Combining the two approaches can improve

your ability to catch market turning points.

TRADING Strategies

BY TERESA LO

From October 1999 until the first weekof January 2000, the QQQ was in a clearuptrend (see Figure 1). On Jan. 5 theuptrend line was breached. The QQQtraded down for one more day before itbegan the bounce to test the high estab-lished on Jan. 3. On Jan. 20 the QQQ

made a marginal high on a test oftop when, combined with thecandlestick made on Jan. 19, itformed a “Harami” reversal pat-tern (see Figure 2).

A Harami reversal pattern is atwo-candle pattern in which thebody of the first candle covers al a rg e r-than-usual pricerange and the second can-dle covers a smaller-than-usual price range. The openand close of the second can-dle must be contained with-in the range of the open andclose of the first candle. Thetwo candles should beopposite colors.

The next day, Jan. 21, wasa very narrow range day,with trading inside therange of the Harami pat-tern. On Jan. 24, a big downday confirmed that the testof the top had failed on thistry.

A few days later, theQQQ made a higher low andbegan a new uptrend that lasteduntil March 14, when a secondshort-term uptrend line wasb reached. A few days later, itbounced to set up the test of theMarch 14 high. On March 23, the

QQQ recorded a marginal new high. OnM a rch 24, it attempted to continueupward but ended the day on a “long-legged doji” pattern, a signal of indeci-sion in the battle between buyers andsellers. The long-legged doji is a candlewhere the upper and lower shadows arevery long, indicating the market wassharply higher and lower during the ses-sion, but the open and close prices arenear the middle of the range covered bythe trading session (see Figure 3).

The next day, March 27, was a narrow-range day where the QQQ traded insidethe range established on March 24, indi-cating more indecision. On March 28, thepattern was broken to the downside,indicating failure on a test of top. Thistime, however, the QQQ started a pat-tern of lower lows and lower highs —the beginning of the Nasdaq downtrend.

For examples of retracements, we willexamine a daily chart of ImmunexCorporation (IMNX) (see Figure 4).

In November and December 1999IMNX entered an uptrend by making aseries of higher highs and higher lowsafter testing the early October 1999 low.After the Dec. 30 high, IMNX retracedfor three days, before making a reversalon a “piercing pattern” on Jan. 5, 2000(see Figure 5).


120.00115.00

110.00105.00

100.00

95.00

90.00

85.00

80.00

75.00

70.00

65.00

60.00

Nov. Dec. Jan. 2000 Feb. Mar. Apr.

Nasdaq 100 Tracking Stock (QQQ)

Test

Test

Candlestick reversal patterns can be used to trade tests of support andresistance levels.

FIGURE 1 TESTS OF TOPS

Source: TradeStation by Omega Research

96

94

92

90

88

90 9⁄16

18 24

Nasdaq 100 Tra c k i n gStock (QQQ)

The body of the second candle is containedwithin the body of the first candle.

FIGURE 2 HARAMI PATTERN


120

118

116

114

112

11027

Nasdaq 100 Index

Long-legged

doji

The market moves sharply higher andlower, but opens and closes near themiddle of the day’s range.

FIGURE 3 LONG-LEGGED DOJI


A piercing pattern is a two-candlereversal pattern that occurs in a down-trend. Both candles should have roughlyequivalent larger-than-usual ranges, andshould have small or no shadows. Thefirst candle opens at or near the high endof the session’s trading range and closesat or near the low end of the session’strading range. The second candle openslower but reverses to close near the highend of the session’s trading range, pro-ducing a white candle.

Between Jan. 21 and Jan. 24, the

December high was tested and then sur-passed; the uptrend continued until Feb.8 when another three-day retracementtook place. On the fourth day, Feb. 14, a“doji” (a candle with a narrow tradingrange, with open and close near thesame price) was formed, indicating inde-cision between buyers and sellers. Thenext day, Feb. 15, was an up day (indi-cated by the white candle), and a three-day reversal pattern called a “morningdoji star” was formed (see Figure 6).

The morning doji star is anotherreversal pattern found indownmoves. The firstcandle has a larger-than-usual price range andopens near the high andthe closes near the low.The second candle is adoji. The third candleopens near the low andcloses near the high.

A moving average canhelp identify the trend,as well as likely reversalpoints. We find that a 30-period weighted movingaverage is useful in iden-tifying short-term trends;retracements in an estab-

lished trend often reverse in the vicinityof the 30-period weighted moving aver-age, alerting the trader to look for can-dlestick reversal patterns. (For moreinformation on moving averages, seeIndicator Insight, Active Trader, June, p.78.)

For example, Figure 4 shows that onceIMNX began trading above the 30-dayweighted moving average in mid-October 1999, the retracements (pull-backs) found support near the movingaverage. Couple that with the series ofhigher highs and higher lows that wereforming, and one would conclude thatIMNX was in an established uptrendand look for spots to enter on each


80.000

70.000

60.000

50.000

40.000

30.000

20.000

Piercingpattern

30-day weighted moving average

Doji star

Oct. Nov. Dec. Jan. 2000 Feb. Mar.

Immunex (IMNX)

Candlestick reversal patterns develop near retracements to the 30-dayweighted moving average.

FIGURE 4 TRADING RETRACEMENTS


A word on order types

I n our trading, we always enterthe market using a stop-limitorder. However, because some

brokers do not allow stop-limitorders, you may need to monitor thestock and, when a certain price isreached, enter a limit order. Ourexit, the protective stop-loss, isalways a stop order.

Stop order: A stop order is anorder that becomes a market order ifand when the market reaches a des-ignated price. A “buy stop” is placedabove the market and becomes amarket order when the securitytrades at or is bid at or above thespecified stop price. A “sell stop” isplaced at a price below the market.It becomes a market order when thesecurity trades or is offered at thestop price or below.

Stop limit order: A stop limitorder is a stop order that becomes alimit order when and if the marketreaches a designated price.

Buy

Stop-loss

After a down day, the stock opens lower but closeshigher, signaling an upside reversal.

FIGURE 5 PIERCING PATTERN


Candlestick chart basics

L ike bar charts, Japanese can-dlestick charts use verticallines to display price action

for a particular time period — a day,a week, an hour, a minute. But can-dlestick charts use shading to high-light the up or down price momen-tum in a given period.

The figure below shows two can-dlestick bars. The high and lowprices of a candlestick are the endsof the vertical line (just like a barchart), but the open and close arehorizontal lines that intersect thevertical line. These two lines arejoined to produce a rectangular areacalled the “real body” of the candle-stick. If the close of the bar is high-er than the open, the real body ishollow, usually white. If the close islower than the open, the real body isshaded, usually black. The portion ofthe candlestick’s trading rangeabove the real body is called theupper shadow. The portion of thetrading range below the real body iscalled the lower shadow.

High High

Low

Open Close

Low

Upper shadow

Lower shadow

Real body

CANDLESTICKS

retracement until the uptrend, alongwith support at the 30-day weightedmoving average, was broken.

Traders often debate the precise namesof diff e rent patterns, but Japanesetraders traditionally place more impor-tance on the real body of the candlestick(the distance between the open and theclose) than the absolute high and low,which are of secondary importance(unless the highs and lows leave verylong “shadows” in relation to the realbody).

With this in mind, wecan safely say the ideal-ized textbook versions ofcandlestick patterns arejust that. Traders need toremember in the re a lworld, the most importantpart of reading candle-sticks is to compare thereal bodies of the candle-sticks to each other.

While it is important tounderstand what a certainpattern means, you alsoneed to know how andwhen to use it. How canJapanese candlesticks helplimit risk? When do weexecute a setup? To illus-trate some rules of thumbfor using candlestickswe’ll look at two buy sig-nals on IMNX.

First, let’s take another look at thepiercing pattern in Figure 5, formed bythe candlesticks plotted on the dailychart of IMNX on Jan. 4 and Jan. 5. Forthe piercing pattern, the important thingis that the first day of the pattern is adown day (a black candlestick). The nextday opens lower but closes near the highof the day, producing a white candle-stick, the real body of which overlapsmost of the black real body from the pre-vious day.

Because IMNX was clearly in anestablished uptrend, traders mightexpect a retracement to be followed by aquick move back to the upside. Becausethe piercing pattern is a reversal signal,we would enter a buy stop limit order onthe break of the high of the second dayof the pattern, in this case the white can-dlestick on Jan. 5, which should also bethe lowest day of the retracement.

If the buy order is not filled withintwo days, it is cancelled on the assump-tion that the power of the buyers takingcontrol at the low is in question and thatanother pattern is forming. If the pierc-ing pattern is invalidated by anotherdownside day, the buy order would notbe filled, we would cancel it, and noharm is done.

If the piercing pattern is confirmed bya move to the upside and the buy orderis filled, a protective stop-loss would beentered just below the low of the day onwhich the trade is executed. In this case,the initial stop-loss order was just below

the low of Jan. 7. On a true retracement,once the uptrend resumes, there shouldbe ample thrust to the upside, as buyerscome back quickly.

While each trader should enter andmove the stop-loss according to his orher own style, it is highly recommendedthat the initial protective stop-loss orderbe placed in the manner described above(rather than at the lowest point of theretracement) for two reasons. First, if theuptrend fails to resume, it may signal adeeper retracement or a change in trend.Second, as so many traders set sell stopsto trigger on a break of the most recentlow, the slippage on protective stop-lossorders in these areas can be significant.

In the next example, we zoom in onthe doji star pattern in Figure 6, formedby the two candlesticks of Feb. 11 andFeb. 14. A white candlestick formed onFeb. 15, confirming a bottom, and withit, a new three-candlestick pattern: the


The white candle confirms the upside reversal fol -lowing the doji candle

FIGURE 6 MORNING DOJI STAR


Stop-loss

Buy

“morning doji star.” The doji star is a two-candle pattern

found both in uptrends and downtre n d s .In an uptrend, the first candle covers al a rg e r-than-usual price range with theopen near the low of the session and theclose near the high of the session. Thesecond candle is a doji. In a downtre n d ,the first candle covers a larg e r- t h a n - u s u a lprice range with the open near the highof the session, the close is near the low ofthe session and the second candle isagain a doji.

These patterns warn the trend may beabout to change and need to be con-firmed by a third candle. For example, amorning doji star reversal pattern in adowntrend consists of a doji star plus athird candle with a close near the high ofa trading session, as shown in Fig. 6.

(The “evening doji star” is a reversal pat-tern found in uptrends and is composedof a doji star plus a third candle with aclose near the low of a trading session.)

There are two ways to trade this pat-tern. Textbooks traditionally cautiontraders to wait for confirmation of areversal pattern before entering a trade.However, it is often advantageous to beaggressive and place a buy stop limitorder at a price where a stock wouldhave to trade on its way to confirmingthe pattern.

In this example, we will use the two-day doji star pattern to set our buy stoporder instead of waiting for the thirdcandle of the three-day morning doji starto confirm that a bottom is in place. Wesimply enter a buy stop above the highof the doji (Feb. 14) itself. If this is the

low for this retracement, a reasonableupside thrust should occur immediately.On Feb. 15, IMNX traded above the highof the doji formed the day before and thebuy stop order was filled.

There are two ways to approach orderentry. If the buy price is triggered, a trad-er can wait to buy at the end of the day,just before the close, after the absolutelow of that day is known. The secondand more aggressive option, the one weuse in our own trading, is to buy imme-diately upon trading at the stop price,and using the intraday low up to thattime as the stop-loss point.

At the end of the day, IMNX closedwell above the high of the doji (in fact, itclosed near the high of the trading day)showing reasonable thrust to the upside.The protective stop-loss would be setjust below the low of the white candle-stick of Feb. 15.

Candlestick patterns are very usefultools because they offer a real-time lookinto the battle between buyers and sell-ers. When added to a trader’s arsenal oftechnical analysis techniques, they offerinsight and help set up trades.Ý


Candlestick patterns are very useful toolsbecause they offer a real-time look into thebattle between buyers and sellers.

A major disad-vantage ofmany techni-cal indicators

is that they only use one pieceof price information — theclosing price. For instance, amoving average or an indica-tor such as Bollinger bands(which plots price bands a cer-tain distance above and belowa moving average) is usuallycalculated using the closingprice — excluding the open,high and low prices. Figure 1shows a 20-day moving aver-age (red line) and 20-dayBollinger bands (green lines)on the Dow Jones trackingstock (DIA), distanced twos t a n d a rd deviations awayfrom the moving average.

One way to work aroundthis is to use the average priceof a price bar — calculated asthe sum of all four price points(open, high, low, close), divid-


116

114

112

110

108

106

104

102

100

98

96Apr. May June July Aug.

DIA-daily

Bollinger bands usually do a good job of containing price action. But because they’rebased on one price observation per bar (the closing price), they don’t provide informa -tion about the “hidden” volatility within each bar.

FIGURE 1 BOLLINGER BANDS


MOVING BEYOND the CLOSING PRICE

BY THOMAS STRIDSMAN

Most indicators have two limitations: They’re usually based on specific price levels rather than price changes, and they only use the closing price of a bar.This technique allows you to build a price-change-based indicator that incorporates all the price points in a bar.

A

MOVING BEYOND the CLOSING PRICE

ADVANCED Strategies

ed by four. Figure 2 shows what this looks like for 5- and 20-day moving averages, respectively. Because the average priceconsists of the different price point observations within eachbar, the 5-day average actually uses 20 different price levels,while the 20-day average uses 80 different price levels.

H o w e v e r, it’s still not possible to make distinct use of the dif-f e rent price points within bars when calculating Bollingerbands. As Figure 2 (compared to Figure 1) shows, there is littled i ff e rence (if any) between Bollinger bands based on the 20-daymoving average calculated using the average price and the stan-d a rd 20-day moving average calculated using the closing price.Also, although there are 20 implicit price points making up the5-day average of the daily average price (blue line in Figure 2),t h e re is only one price observation per bar (the average price)going into the Bollinger band calculation. To make the standarddeviation calculations for the Bollinger bands statistically re l i-able using traditional statistical guidelines, however, we shoulduse at least 20 (or 30 or more, depending on the source) explicitprice observations in our calculations.

A second disadvantage of most indicators is that they usual-ly are calculated using the actual price levels instead of theprice changes from bar to bar (or some other period). Thismakes little sense because price levels are likely to change overtime as a market trends up and down, while the percentagechanges from one period to the next are more likely to stay thesame. Basing indicator calculations on price changes ratherthan actual prices makes it easier to compare indicator read-ings, both between different markets and different time peri-ods.

One way to calculate a movingaverage on price changes ratherthan actual prices is to measurethe average price change over acertain time period and thenadd that change to the closingprice (or the average price) ofthe most recent bar. For exam-ple, you could calculate theaverage day-to-day pricechange for the last five days,and add that to the most recentclose.

Figure 3A shows what thislooks like. It’s a 20-day movingaverage of close-to-close p e r-centage changes added to theclosing price of the second-to-

last bar. This line shows where the market would have closedon the following bar (the most recent bar in the chart, which isnot a part of the calculation) had it followed the average rate ofchange of the last 20 days.

Moving this calculation forward one day provides an indi-cation of where the market should close tomorrow. This adjust-ed indicator, which is a 20-day moving average of the percent-age price changes, is shown in Figure 3B. Notice how it differsfrom the regular 20-day moving average (blue line), whichreally doesn’t tell you anything about where the market isgoing in the short term.

For TradeStation users out there, the EasyLanguage code is:

Inputs: LookBack(20);Variable: PercChange(0),

NewLevel(0),AvgPercChange(0);

PercChange = (Close - Close[1])/Close[1];

AvgPercChange = Average(PercChange,20);

NewLevel = Close * (1 + AvgPercChange);Plot1[1](NewLevel,””);

To calculate and plot the standard deviation bands aroundthis new moving average (creating a Bollinger band based onthe changes in the closing price), add the following code:


116

114

112

110

108

106

104

102

100

98

96

Apr. May June July Aug.

DIA-daily

Bollinger bands calculatedusing average price

5- and 20-daymoving averages

calculated on average price of bar.

Basing calculations on the average price (as was done here) might provide a betterfeel for what happened at each particular bar, but it still won’t let you make directuse of all price points, rendering this approach no more useful than using closingprices only.

FIGURE 2 USING THE AVERAGE PRICE


Inputs: StDevs(1);Variable:StdDevChange(0),

U p L e v e l ( 0 ) ,D o w n L e v e l ( 0 ) ;StdDevChange =

StdDev(PercChange, 20);UpLevel = Close *

(1 + AvgPercChange + StDevs * StdDevChange);

DownLevel = Close * (1 + AvgPercChange - StDevs * StdDevChange);Plot2[-1](UpLevel,””);Plot3[-

1](DownLevel,””);

F i g u re 4 shows the re s u l t .The upper and lower bandsnow are the Bollinger bands ofthe change in the closing pricerather than the closing priceitself. If you compare theset h ree lines with the ones inFigure 1 it’s obvious which onesmore closely follow price and,therefore, are most useful to ashort-term trader.

Making a statistical indicatorH o w e v e r, this indicator stilldoesn’t address the main prob-lem of not incorporating thehigh, low and opening prices.To do this, the calculations get abit more complicated. What weneed to do is create an array (aset of sequential values) hold-ing the necessary price data. Foreach new bar, the four oldestprice points (open, high, lowand close) will be thrown out ofthe array and substituted withthe corresponding price pointsfor the most recent bar. Forexample, to calculate a movingaverage consisting of 20 pricepoints, you will need an arraycontaining the price data (open,high, low, close) of five pricebars.

To just use the actual pricelevels, however, won’t makethis indicator any different thanthe 5-day moving average cal-culated on the average price ofthe bar (the blue line in Figure2). To make the indicator more


114

113

112

111

110

109

108

107

106

105

104

103

102

May June July Aug.

DIA-daily

20-day moving average of close-to-close price changes,

added to closing price

Basing calculations on price changes rather than price itself creates an indicator thathugs the most recent price action more closely, making it more meaningful for short-term traders. Figure 3B compares a 20-day moving average of price changes (red line)to a moving average of the actual price (blue line). The blue line only tells you whatthe average price of the last 20 days is; the red line tells you where the market islikely to close the next bar given the average change in prices over the same period.

FIGURE 3A USING PRICE CHANGES

110

109

108

107

106

105

17 24 31 Aug. 7

DIA-daily

FIGURE 3B

Source for both: TradeStation by Omega Research

20-day moving average of price changes

Standard 20-day moving average

useful for short-term trading,look instead at the price changesfrom the previous bar’s close,just as we did for the indicatorsin Figures 3B and 4. In otherwords, calculate the differencebetween yesterday’s close andtoday’s open, high, low, andclose, respectively. For instance,to calculate the perc e n t a g echange between yesterd a y ’ sclosing price and today’s high,the EasyLanguage formula is:

HighPercChange = (High -Close[1])/Close[1];

In a more generic form theEasyLanguage code for thee n t i re indicator, including the20-price-point moving averageand its upper and lower stan-d a rd deviation boundaries,would look something like this:

Input: VSStd(1);Vars: SumVS(0), AvgVS(0),DiffVS(0), StdVS(0), SetArr(0),

SumArr(0), DiffArr(0), VSLow(0), VSMid(0), VSHigh(0);{First we define the array}

Array: VS[20](0);

{Then we’re using a loop function to fill it with the d i f f e r e n tprice changes} For SetArr = 0 To 4 Begin

VS[SetArr * 4 + 0] = (O[SetArr] - C[SetArr + 1]) / C[SetArr + 1];VS[SetArr * 4 + 1] = (H[SetArr] - C[SetArr + 1]) / C[SetArr + 1];VS[SetArr * 4 + 2] = (L[SetArr] - C[SetArr + 1]) / C[SetArr + 1];VS[SetArr * 4 + 3] = (C[SetArr] - C[SetArr + 1]) / C[SetArr + 1];

E n d ;For SumArr = 0 To 19 Begin

If SumArr = 0 ThenSumVS = 0;

SumVS = SumVS + VS[SumArr];If SumArr = 19 Then

{Here we calculate the average price change over the period}

AvgVS = SumVS / 20;For DiffArr = 0 To 19 Begin

If DiffArr = 0 ThenDiffVS = 0;

{Then we calculate the standard devi a t i o n }DiffVS = DiffVS + Square(VS[DiffArr] - Av g V S ) ;If DiffArr = 19 Then

StdVS = SquareRoot(DiffVS / 20);E n d ;

E n d ;{ F i n a l l y, we add the moving average (and the standard devia -tions) to the latest close for an indication of tomorrow’s trad -ing range}

VSLow = C * (1 + (AvgVS - StdVS * VSStd));VSMid = C * (1 + Av g V S ) ;VSHigh = C * (1 + (AvgVS + StdVS * VSStd));P l o t 1 [ - 1 ] ( V S L o w, “VS Low”);Plot2[-1](VSMid, “VS Mid”);Plot3[-1](VSHigh, “VS High”);

Figure 5 shows what this indicator looks like. Because everyindicator needs a name, we hereby dub this the MeanderIndicator, and the standard deviation boundaries the Upperand Lower Meander Boundaries.

To get a feel for how this indicator works compare it tomeasuring the weight of four different persons you meetthroughout the day. The first and the last ones you meet couldbe your opening and closing weights, respectively. The othertwo you pick at random sometime during the day. After fivedays, when you have met 20 people, you can calculate theiraverage weight and the standard deviation boundaries aroundtheir average weight.

Say on the next day (the sixth day) you go out and weigh thefirst person you meet, and it so happens that he or she is skinand bones, and well below the lower standard deviation boun-dary. As a result, your statistical calculations tell you the otherthree persons you’ll meet that day will likely weigh more.

When the sixth day is over, you add that data to the array


110.5

110

109.5

109

108.5

108

107.5

107

106.5

106

105.5

105

104.5

17 24 31 Aug. 7

DIA-daily

Calculating the Bollinger bands of price changes creates the upper and lower standarddeviation boundaries for where the market is likely to close the next bar. If the bandsare plotted one standard deviation boundary away from the average, they can beexpected to contain 67 percent of the price action. If they’re two standard deviationsaway from the average, they should contain 95 percent of the price action.

FIGURE 4 MEANDER INDICATOR


and throw away the data fro mfive days ago. This is akin tomaking sure that you alwaysuse the most accurate, up-to-date data, which you’d need ifyou were collecting weight dataa round in the world. For exam-ple, if you live somewhere in theMidwest, the average weight ofthe people you meet pro b a b l ywould be higher than that of thepeople you meet on vacation inAsia (where people generallyhave a smaller build). Carryingout this line of thought, it also isa well-known fact that peoplee v e r y w h e re are getting heavier,so to keep your calculations asaccurate and up-to-date as pos-sible, you need to get rid of theo l d e r, obsolete data.

How to use the indicator: Estimating price movementS i m i l a r l y, when trading themarket using this indicator, ifthe opening price is below thelower standard deviationboundary, you know that thelow price of the day also will belower than the standard deviation boundary. Consequently,there is a good chance that: 1) at least the opening price will notequal the high price of the day; and 2) the closing price also hasa good chance to be higher than the open.

On the other hand, if the opening price happens to fall some-w h e re between the upper and lower standard deviationboundaries, but the market subsequently tests any of these lev-els, chances are pretty good that the closing price also will fallsomewhere between the boundaries.

If you set the boundaries one standard deviation away fromthe average, the laws of statistics say that approximately 67percent of the daily price action will take place within theboundaries. If you set the boundaries two standard deviationsaway from the average, you can expect them to contain

approximately 95 percent of all the price action. Remember,though, this won’t always be the case and the price actionwithin each bar won’t necessarily be equally distributedaround these levels.

As you can see from Figure 6, every now and then there willbe a bar that is completely outside of the boundaries, but witha little experimentation you should be able to come up with amoving-average length and standard-deviation distance thatsuit your particular trading style, and that help you get a feelfor when the market is most likely to reverse and trade higheror lower.

To copy the programming code used in this article, see the online ver -sion at Active Tr a d e r’s Web site www. a c t i v e t r a d e r m a g . c o m .

Ý


1 1 0 . 5

1 1 0

1 0 9 . 5

1 0 9

1 0 8 . 5

1 0 8

1 0 7 . 5

1 0 7

1 0 6 . 5

1 0 6

1 0 5 . 5

1 0 5

1 0 4 . 5

17 24 31 Aug. 7

DIA-daily

Basing calculations on the percentage changes between one bar and the close of theprevious bar allows you to shorten the lookback period to only five bars, providing aneven better indication of what might happen the next bar. Depending on where themarket opens and the subsequent price action, you can use the upper and lower“Meander” bands as triggers for limit-order trades.

FIGURE 5 SHORT-TERM PERSPECTIVE


Basing indicator calculations on price changes rather than actualprices make it easier to compare indicator readings, both

between different markets and different time periods.

BY MARK ETZKORN

F or Mark McDonnell, the hours are right for trading.The 43-year-old former engineer spends about anhour a day trading — and he’s usually only in themarket for a few minutes.

Texas-based McDonnell has been trading full-time for thelast year-and-a-half after spending two-and-a-half years atFidelity Investments — a job he took after leaving engineering.Like many traders, his first exposure to the markets wasthrough mutual funds. But in addition to dollar-cost averaginginvesting, he actively traded funds, analyzing their perform-ance over different time periods to find those with the highestlikelihood of upside movement. When he moved on to Fidelity,he began blending options into his trading.

The switch to full-time trading came after he made animportant discovery.

“I was making more money selling call options than I wasworking,” he says. “So I asked myself, ‘Why do I need to dothis?’ I also took some profits in the real estate market, and thatwas enough backup money to go out on my own.”

His early approach was to write covered call options (sellingcalls against long stock positions), a strategy that was prof-itable for a while. However, McDonnell went through a roughpatch last year that brought home the risks of the strategy. Hecontinued to sell additional options if the stock kept dropping.

“The main problem with selling calls is you still have therisk of stock ownership,” he says. “The best scenarios occurwhen the stock goes up or stays even. I had problems havingto sell some stocks that looked like they were breaking downtechnically. All the calls in the world won’t help you if yourstock falls apart.

“The good thing was that I became very comfortable withselling options. You can make 5 to 10 percent monthly, conserv-a t i v e l y, on the premium. You just have to make sure none ofyour stocks get away from you. You have to watch them — trackyour intraday highs and lows to see if they’re breaking down.”

McDonnell traded mostly lower-priced stocks (around $25or less) with his call-writing strategy, because, as he points out,“If you have a stock trading at $101, the 100 calls have anintrinsic value of 1. If you trade a $21 stock, the 20 calls stillhave an intrinsic value of 1.”

His current trading approach, which he began to developwhile still at Fidelity, consists of looking for quick rebounds inl a rge-cap, institutional stocks that have suff e red sharpdeclines. He buys if there’s evidence of some upside momen-tum (using a program called WizeTrade) as the stock is closing,and exits the stock the following morning on the open.

“What I would do is trade the ‘big boys’ — the TXNs (TexasI n s t ruments) and SUNWs (Sun Microsystems),” he says.“When the Nasdaq was down 100 to 150 points, I would lookfor the stocks that were getting killed. I would write down theintraday lows. If they broke their intraday lows near the closeon increasing volume, but were still off (higher than) their newlows, I’d buy around 300 shares on the close. These stockswould gap up the next morning, and I’d sell them right on theopen. Let’s say SUNW is down 10 points for the day, and gapsup two points, I’d take my $600 profit.”

McDonnell, who says this approach has worked for himaround 80 percent of the time, uses options when a stock opensagainst him the next morning. “If the market opens flat, youcan always scratch the trade,” he says. “If it gaps down, say,two points the next morning, you can sell calls.”

For this contingency plan, McDonnell sells slightly out-of-the-


The overnight O P T I O N

Trading set-upHardware: Pentium 700 PC, 20 GB hard drive,

128 MB RAM, 17-inch monitor

Internet connection: ADSL (1.5 million bytes/sec)

Brokerage: Fidelity (www.fidelity.com)

Software: WizeTrade (www.wizetrade.com), for price/trend analysis

Data: PCQuote.com (www.pcquote.com), Nasdaq Level I data

“ I’m not greedy. Even if some of the signals are still green, I’m not going to try to squeeze it.”

The Face of TRADING

money calls, pointing out that the premiums onthe options will be inflated because of the sharpd rop the stock just experienced.

“Remember, I’m buying a stock that was down10 points yesterday,” he says. “That drives thepremiums up. You’re going to get a premium offive or six points on that kind of option, andyou’re only down two points on the stock.”

In situations where the stock continues todrop, McDonnell will look to get out of his stockwhen the loss is equal to (or slightly less than) the premium hecollected on the option, scratching the trade or taking a smallprofit. “Then I have to watch the stock, because I may have tocover the calls if the market goes back up,” he says.

McDonnell estimates the risk on his overnight trades is, onaverage, six to seven percent of his total account equity; therest is in cash, earning interest. His market exposure is oftentwo minutes or less: He will frequently buy in the last 30 sec-onds of trading and sell immediately at the open.

He says the fact that he’s trading large-cap stocks with heavyinstitutional interest also works in his favor. “What’s wro n gwith having 300 shares of the big boys in your account?” he asks.“I would always look at the tech-stock mutual funds’ biggestholdings and go with those stocks, because the odds are betterthat you’ll be OK in the long-term. If the stock stays down twopoints for another month, I can make more money selling calls.”

While some traders prefer to find stocks that trade stro n g l y

into the close as candidates for next-morning gap plays, McDonnell considershis approach to be “playing it safe. I’msticking with the big boys, waiting forthem to get hit hard, and looking forthem to bounce back.”

Arepresentative trade of McDonnell’swas the gap opening on Tr a n s o c e a nSedco Forex (RIG) on Aug. 21 (see Figure1). The stock gapped lower on Aug. 20from the previous close of 57 11⁄16 andestablished an intraday low of 52 5⁄8around 12:45 p.m. EST. The stock then

moved sideways to slightly higher for the remainder of the ses-sion, closing at 53 1⁄2 on increasing volume. McDonnell boughtat the close. On Aug. 21, the stock gapped open higher at 54 3⁄16

and immediately moved up. McDonnell sold out for a 1 1⁄4-point profit. (Two other stocks on his list the same day wereE fficient Networks (EFNT) and Advanced Micro Devices(AMD); both gapped open higher as well.)

McDonnell might also make a couple of day trades perweek. A typical trade is something of an inversion of hisovernight approach. He’ll find stocks that are up significantlyon the day (“10 points or more on a stock trading at less than$100”), wait for an afternoon lull, then buy if they show upsidemomentum in anticipation of a rally into the close.

But the overnight approach remains his bre a d - a n d - b u t t e r. “IfI can start looking for stocks an hour before the close, and sellthem the next opening, I’m working an hour a day,” he notes.

McDonnell likes the low-impact approach to trading andwarns against over-trading. He describeshyperactive, intraday scalping as “insani-t y. I could never do it — I don’t want to doit. With that type of stress, just turn thecomputer off and get a real job. You caneither work hard or work smart. The re a-son I like this overnight method is it’s ah i g h - p e rcentage approach, and I have thecontingency plan of selling calls if thestock opens against me.”

Steady returns are McDonnell’s goal —not swing-for-the-fences monster trades.“I’m not gre e d y,” he says. “If I get a point,I’m out. Even if some of the signals arestill green, I’m not going to try to squeezeit — I’m not going to risk my profit for thed a y. I’m very conservative.

“I’m trying to pull in about $400 ad a y,” he continues. “There’s got to besomething there every day. I have a sys-tem with a backup plan, and my markete x p o s u re during the day is very low. ”

His experiences last year taught himthe value of flexibility. “It’s a learningprocess, and hopefully we’re all gettingsmarter,” he says. “It hasn’t all been pen-nies from heaven, but I’ve learned mylessons. I’m putting money in my pocketeach day.”Ý


RIG, two-minute chart, with daily chart insert. The stock sold off sharply thefirst day, consolidated, and closed on increasing volume. A long position wasestablished on the close and exited the following morning on the gap-up opening.

FIGURE 1 THE OVERNIGHT APPROACH

Transocean Sedco Forex Inc.(RIG), 2-minute

58

57

56

54

54

53

56

52

48

Aug.24 31 7 14 21

9:30 10:00 10:30 11:00 11:30 12:00 12:30 13:00 13:30 14:00 14:30 15:00 15:30 9:30 10:00 8/21 Monday 8/22 Tu e s d a y

Stock gaps lower,sells off

Increasing volumegoing into close

Stock gaps higher

Volume


2 0 0 , 0 0 0

1 0 0 , 0 0 0

55

“ You can either work hard

or work smart.”

AT: What are some of the most commonweaknesses you see in traders?VT: I’ve designed a test to help peopledetermine their weaknesses. There are10 measured areas that fall into threecategories: psychology and discipline;understanding systems and expectancy;and money management, which I nowcall “Position Sizing.”

I think you could safely say that every-one who wants to succeed as a tradermust master those three areas. However,all of them fall under psychology becausewe are all human beings, not ro b o t s .Psychology impacts everything.

One reason psychology is so impor-tant is that we all have certain biasesthat help us deal with processing thevast amount of information to which weare exposed everyday as traders. Thesebiases help in decision-making, but theytend to lead us astray. If you naturallyfollow those biases — and most peopledo — then you will have a very hardtime learning what you need to learn tobe successful. And incidentally, most ofthe information available to help traderscaters to our biases.

QAConquering trading biases:

While many traders waste time chasing elusive “perfect” entry techniques, onetrading coach points out that trading success is about accepting market realities,knowing yourself and implementing a money management plan that keeps theodds on your side.

Van K. Tharp. Ph.D., is a well-known trading coach and

consultant. He is president of the International Institute

of Trading Mastery (www.iitm.com) in Cary, N.C., and

author of “Trade Your Way to Financial Freedom” (1999, McGraw-

Hill) and “Financial Freedom Through Electronic Day Trading,” due

to be released this fall.

He recently spoke with Active Trader about the “psychological”

biases that prevent traders from succeeding in the markets, and what

traders can do to combat negative habits.

“When you feel like you have

to be right, the result is that you

won’t take losses.”

QA& w i t h DR. VAN THARP

TRADIING Psychology


to risk more after a few losses and risk less after a few wins. But the best way to tradeand make money is to always risk more when you are ahead.

Last, people seem to think they know what they’re doing and none of this — theissues I’m talking about now — applies to them. They don’t take responsibility for theirresults and they blame others for what happens to them. When you do that, you willnever learn from your mistakes. Instead, you’ll continually repeat them. And if yourisk a constant percent of your equity, you’ll be doing just that.

I think that if people could master, really master, those four concepts, they’dbecome good traders. The problem is that they’re not as easy to master as theyappear to be.

AT: What do you consider the hallmarks, psychological or otherwise, of a goodtrader?VT: I teach people that there are seven key principles you must master to be a goodtrader. The first is what I was just talking about — you must take total responsibilityfor what happens to you. Whatever your results, you must believe that you somehowproduced them. This means that when things go bad you can start to learn from yourmistakes. It also means that most of your effort will be in the area of understandingyourself, not understanding the markets.

Second, you must treat your trading like a business. You must know who you are ,what your mission is as a trader and have objectives that fit your mission. Once youhave those things in hand, you can then design a game plan that will help you succeed.

Third, you must understand that whenever you enter a trade you must have somepoint at which you will get out in order to preserve your capital. I call this your ini-tial risk level, or “R” for short. If you don’t have an exit point when you enter a trade,then you are not a professional and you have very little chance of long-term success.

Fourth, you must strive to make profits that are multiples of R. For example, ifyour profits are generally 10R, then you can be wrong 70 percent of the time and stillmake tremendous profits. For example, seven 1R losses and three 10R gains wouldgive you a total profit of 23R. That’s superb. When you do this, you will have a “pos-itive expectancy” system.

Fifth, you must understand a what a low-risk idea really is: a trading approachwith a positive expectancy, but traded at an appropriate size to allow for the worstpossible outcome in the short run so that you can realize the expectancy over time.

Sixth, you must master Position Sizing, which is the part of your system thatanswers the question, “How much?” throughout the course of the trade. It is the keyto meeting your objectives and it is the most overlooked aspect of trading, with theexception of psychology.

Finally, you must continually work on yourself, because your trading results aremuch more a result of you and what you do than any other factor. I have a PeakPerformance Course for traders that addresses many of these issues like discipline,the 10 tasks of successful trading, self-sabotage, smart decision making, etc.

AT: Do you think trading skills can ultimately be taught or does a basic aptitudefor the profession play a bigger part in determining success?VT: It’s hard for me to imagine anyone who masters those seven principles not beinga success. But few people can do it. First, most people are not ready for principle No.1, and unless you master that one, you have little chance of success, in my opinion.

Also, it’s very hard to understand some of the other principles unless you havegood mathematical skills. For example, concepts like expectancy, which comes fromprinciples No. 3 and No. 4, are very difficult to understand if you don’t have goodmathematical stills. And math skills are critical for mastering Position Sizing.

I give a simple 1 1⁄2-hour talk in which I teach the essentials of expectancy and positionsizing. However, I’ve had people listen to that talk, think it was brilliant and still not getthe concept. Then they want to take our most advanced seminars before they master thefundamentals. That never works, but that’s the way most people want to go.

I also think commitment is critically important to be a good trader. If you have com-mitment you can overcome major obstacles. If you don’t have it, nothing can help you.

AT: If you could only communicate one principle, rule or guideline to increase the


AT: What are some examples of thesebiases?V T: T h e re are four that actually coverthe common weaknesses almost alltraders have.

First, we all need to be right. We’retaught in the school system that any-thing less than 70 percent is failure. As aresult, we want to be right at least 70 per-cent of the time in the markets. But whenyou have this bias, the result is that youwon’t take losses. You’ll think the losingtrade you’re holding might turn aroundand become a winner, so you won’t takethe loss. Eventually the loss becomeshuge so you have to take it or you turninto a long-term investor.

The flipside of having this bias is thatyou want to take profits right away.Why? If you take profits right away,your trades have no chance of turninginto losers and you get to be “right.”

But this flies in the face of the goldenrule of trading: “Cut your losses short andlet your profits run.” By needing to beright you end up doing the opposite, cut-ting profits short and letting losses ru n .

Second, we want to be in control,which means, in trading, we want tocontrol the markets. The only thing mostpeople can control about the markets iswhen they get into a trade.Consequently, most people focus all oftheir effort on predicting the market andpicking the right stocks.

But I think most of this effort is wast-ed. Profits come from exits, not pickingthe right stocks. If you cut losses and letprofits run, you can be a phenomenalsuccess — and it will have little to dowith what stocks you pick.

Third, people don’t understand thenature of streaks. A random sequencetends to have very long streaks in it —longer than most people would think.But since we don’t expect them, we tend

“The key to success is to find

a trading approachthat fits you.”

odds of trading profitably, what would it be?VT: The one that’s hardest to learn — taking responsibility for whatever happens toyou. When you’ve mastered that, then you are in total charge of your life. Untilyou’ve mastered that, you’re a helpless victim. But as I’ve said, few people want toclaim their power. They want “know-how,” but they don’t want self-mastery.

AT: Do you feel a systematic trading approach is vital to success?VT: Yes and no. We have programs designed to teach people to be professionaltraders. In that program, I require them to develop a systematic game plan that willmake at least 50 percent per year with very few drawdowns. That’s actually quiteeasy to do when you’ve mastered the seven principles, as long as you are not trad-ing too much money — say, $5 million or less.

I n i t i a l l y, it must be very mechanical. But later, when people have proven they cantrade mechanically, I think they become much better by learning how to break the ru l e s .

AT: How important is money management in your overall philosophy? VT: I believe it’s the key to meeting your objectives in the market. What I now callPosition Sizing is the part of your system that controls how much you trade through-out the course of a trade.

In my workshops we play a simple game that involves randomly drawing marblesout of bag and replacing them. Let’s say 60 percent of them are winners and 40 per-cent of them are losers — that’s not a bad system. People start out with $100,000 andthey decided how much to risk on each marble draw.

After about 50 draws from the bag, with everyone in the room getting the samemarble draws, we’ll generally have as many different account levels as there are peo-ple in the room, with the exception of the people who go bankrupt. There are onlytwo factors involved in that game — Position Sizing and psychology.

I now have a five-part computer game that allows people to master PositionSizing. It starts out with a simple system like the one described and it progressesthrough five increasingly more complex areas to help people master Position Sizing.

AT: What is the core money-management concept or technique you teach? VT: Risk a percentage of your equity. And for stock traders, I think 1 percent is high— one-half percent might be a better starting point.

AT: How would you rank the following aspects of trading in terms of importance:entry rules, exit rules, money management/risk control, psychology and capital -ization?VT: Well, psychology is definitely No. 1 because we are human beings and every-thing we do from money management to entry rules requires us to use our brains.

Someone once told me that psychology wasn’t important to his firm because theywere totally computerized. That firm later went out of business because his partnerdidn’t take a particular trade because they had so many losses in that particular item.That trade would have been the trade of the year and saved their business. The part-ner didn’t take the trade because of his psychology — and that one incident costthem the business.

Money management/risk control would be No. 2. By this I mean Position Sizingas I’ve described it and defining what multiple of your risk level you use on eachtrade. This is fundamental. Without it, no trader can succeed.

Capitalization is No. 3 because without adequate capital you cannot practice adequatePosition Sizing. Your risk will always be too big and you may eventually blow out.

Exit rules are No. 4. Exit rules are what allow you to cut your losses short and letyour profits run.

Entry rules are the least important. They are where most people put their empha-sis because they believe that prediction and finding the right stock are so important.Pick some stocks that have the power for a big move and you’ll be fine if you havethe other principles mastered.

AT: Have you found there is a particular kind (or general category) of trading strat -egy or approach that successful traders tend to use?


VT: No, everyone is different. The key tosuccess is to find something that fitsyou. That’s really the holy grail of trad-ing. However, if you don’t know whoyou are, then how can you find some-thing that fits you?

AT: What are the most common pitfallsfor beginning traders?VT: They ask the wrong questions, suchas: What should I buy and what is themarket going to do? The seven princi-ples I outlined earlier have little to dowith these questions. Unfortunately, thebeginning trader has trouble movingbeyond them.

AT: How can traders reinforce goodhabits and break bad ones?VT: I designed a model for good tradingthat incorporates 10 tasks. Three ofthem are designed to reinforce goodhabits and break bad ones.

The first task is self-analysis. Yourtrading performance is a function ofyou, so you must spend time every dayanalyzing yourself.

The second key task is mentalrehearsal. You must decide what couldhappen today to cause you to breakyour rules. Mentally rehearse how youwill deal with those situations, so whenand if they come up, your response tothem will be automatic.

The third key task is a daily debrief-ing. At the end of the day, ask yourself,“Did I follow my rules?” If you did,then pat yourself on the back — even ifyou lost money. If you didn’t, then youneed to figure out why and use themental rehearsal step to correct the mis-takes in the future.

And here’s the most important aspectof the daily debriefing — if you don’thave any rules to guide you, then youare really in trouble and you shouldn’tbe trading.

**”Position Sizing” is a trademark of Position SizingTechnologies Inc.

Ý

“ Profits come from exits, not picking the right stocks.”

BY MIKE CHALEK

A famous trader in the early1900s described technicalanalysis as “the study offactors arising in the mar-

ket itself which can influence pricemovement without regard to fundamen-tal considerations.”

Technical trading systems reflect thisby using a set of specific rules that gen-erate a set of entry and exit signals forthe trader to act upon, no matter whatthe fundamental circumstances mightlook like.

Analyzing entry and exit signals inrelation to the outcome of trades helpsdetermine the true effectiveness and effi-ciency of a trading system. The rules of aspecific system are not importantbecause we simply are trying to under-stand the characteristics of the differenttrades in order to perform a kind ofstrategy “tune-up.”

This type of analysis, known asMaximum Adverse Excursion (MAE)and Maximum Favorable Excursion(MFE), was popularized by JohnSweeney in his books Campaign Trading:Tactics and Strategies to Exploit the Markets(John Wiley & Sons, 1996) and MaximumAdverse Excursion: Analyzing PriceFluctuations for Trading Management

(John Wiley & Sons, 1997). All you needto calculate MAE/MFE are the entrypoints for each trade, the subsequentprice action within the trades and their

individual exit points.The MAE level for each trade is

defined as the lowest low of the openp rofit within the trade or, stated diff e re n t-l y, the maximum open loss within thetrade (which would be zero if the trade isnever in negative territory). The MFE


Less is M O R E

Entering the market

is the easy part.

But once you’re in,

you’ll need to figure out

when to exit or add

to your position

to maximize your returns

in the long run.

RISK Control and MONEY Management

M O R E

70

60

50

40

30

20

10

0

234

5$1,000$2,500$2,000$1,500

$3,000$4,500$4,000

$5,000$3,500

You can get a good indication of whether momentum is working for a tradeand whether you should add to the trade by measuring the likelihood ofmeeting different profit targets after a trade has lasted a certain number of days.

FIGURE 1 PROFIT TARGET LIKELIHOOD

Daysintrade

Profit target

level is defined as that point within thetrade where the open profit is at its peakv a l u e .

Many of us are more or less concernedwith our trading systems when it comesto excursion analysis, especially if wealready have a system that makes usmoney, but also constantly keeps us onedge because of large price swings.Analyzing the low or negative extremesof the price movements within eachtrade (the MAE) provides a good indica-tion of where to place your initial stop-loss or trailing stop orders. But did youalso know you could increase profits anddecrease drawdown by optimizing yourexit techniques and maybe even cuttingthe winners short as well?

This is the MFE part of the MAE/MFEanalysis. It can be done without chang-ing the system’s rules or adding anymore variables. For this article, we willuse the Portfolio Maximizer softwarepackage that was co-developed byOmega Research Inc. (www. o m e g are s e a rch.com) and Rina Systems Inc.( w w w.rinasystems.com). Rina Systemsconsiders MFE to have support andresistance characteristics in much thesame manner as regular price charts.

The theory is that when price pene-trates a specific resistance level, thatresistance level becomes a future sup-port level. The same concept can beapplied once an open profit has pene-

trated a specific dollar or percentagelevel, which also is just another way ofmeasuring the momentum of the trade(as opposed to using a traditionalmomentum oscillator). When the openprofit penetrates a resistance level, thetrade typically remains above that levelfor the remainder of the trade. The goalof this strategy, once the resistance levelbecomes identified, is to add to existingpositions to improve total performance.

After all, the open equity of each tradewill always be a perfect image of the cor-responding price moves in the market,but on a different scale. So, instead ofanalyzing support and resistance on aprice chart (not knowing if it would berelevant because we don’t knowwhether or not we would have been in atrade at that particular time), we insteadanalyze support and resistance on theopen equity of each trade, where it willbe of more importance.

An alternative approach is to examinewhat happens to a trade a certain num-ber of days after it has been entered. Thefollowing example will look at ways toimprove a short-term, trend-following,stop-and-reverse system that is alwaysin the market, either by adding to theoriginal trade at a specific day or by tak -ing advantage of any possible profit tar-gets before the system reverses on thenext signal. We will call this type ofanalysis TBE or Time-Based Excursion

analysis. The market traded is the S&P500 index futures market from January1999 through June 2000. Note: The fol-lowing numbers don’t say anythingabout the end result of the trade, butonly indicate the likelihood for a certainprofit target to be reached or not.

Figure 1 shows that on the second dayafter the original entry signal, there is a67 percent chance that the trade willreach an open profit of $1,000, whichgives it a mathematical expectancy of$670 (0.67 * 1,000). There is a 19 percentchance it will reach an open profit of$5,000, for a mathematical expectancy of$950 (0.19 * 5,000).

However, this analysis says nothingabout what happens to all the losingtrades. Table 1 shows, with a $1,000 prof-it target, of all trades still open on thesecond day, the winners amounted to$180,000, while the total amount lost forall those trades that didn’t reach theprofit target amounted to -$137,500.For the $5,000 profit target, the samenumbers come out to $260,000 and -$264,000, respectively, thus rendering anet loss of $4,000. This shows that youcannot look at profits alone when decid-ing where to place your profit target.

Note, however, that for the furthestrow of columns in Figure 1, which repre-sents the likelihood for a certain profittarget to be reached after the fifth day ofthe original entry, there is a little hump


Not all trades will reach the specified profit target. The ones that don’t will end up as losers. The day-to-day changebetween the winners and the losers for a specific profit target will help you identify shifts in momentum and when itcould be a good idea to add to a trade.

TABLE 1 GOOD TRADES, BAD TRADES

Ta r g e t s Day 2 (Win) Day 2 (Loss) Day 3 (Win) Day 3 (Loss) Day 4 (Win) Day 4 (Loss) Day 5 (Win) Day 5 (Loss)

$1,000 $180,000 -$137,500 $101,000 -$78,000 $40,000 -$40,000 $20,000 -$15,000

$1,500 $217,000 -$179,000 $115,000 -$114,000 $51,000 -$50,000 $24,000 -$25,000

$2,000 $252,000 -$204,000 $132,000 -$133,000 $56,000 -$55,000 $28,000 -$26,000

$2,500 $267,000 -$221,000 $140,000 -$154,000 $60,000 -$70,000 $35,000 -$26,000

$3,000 $276,000 -$237,000 $156,000 -$170,000 $69,000 -$73,000 $39,000 -$26,000

$3,500 $276,000 -$249,000 $150,000 -$178,000 $77,000 -$73,000 $42,000 -$30,000

$4,000 $260,000 -$254,000 $156,000 -$191,000 $76,000 -$81,000 $36,000 -$36,000

$4,500 $261,000 -$257,000 $157,000 -$200,000 $67,000 -$86,000 $27,000 -$44,000

$5,000 $260,000 -$264,000 $150,000 -$203,000 $70,000 -$93,000 $30,000 -$45,000

between the profit target levels $2,000and $3,500. This indicates the mathemat-ical expectations for these levels are evenhigher when the trade has lasted at leastfive days.

To investigate this further, you can

look at Figures 2a and 2b, which showthe differences between the gross profitsand the gross losses, in dollars and per-centages, respectively. Note that for days3 and 4 it doesn’t seem to matter whereyou place the profit target —the overall

result for the system will be a net lossanyway. But at the fifth day of the trade,something interesting happens.Suddenly the profit targets in the $2,000to $3,500 region become net profitableagain. It is almost as if the market, afteran initial pullback and some hesitationafter the entry, confirms that the originalentry signal was valid and it now, there-fore, could be worthwhile to add to thetrade (if it is, in fact, profitable).

Yet another application of the MFEconcept is to look at the maximum openp rofit of the trade in comparison to itsfinal profit. Trend-following bre a k o u ttrades tend to rack up good profits onlyto give back a large portion of those pro f-its before the system signals an exit.F i g u re 3 shows the relationship betweeneach trade’s open profit and its final out-come. The horizontal axis re p resents themaximum open profit for each trade,while the vertical axis re p resents theclosed profit or loss for each individualtrade. Both winning and losing trades areplotted on the same graph. The green tri-angles, pointing upward, re p resent win-ning trades and the red triangles, point-ing downward, re p resent losing trades.

F rom Figure 3 you can determine on atrade-by-trade basis just how efficient thesystem is. For example, the most idealtrade would be a winner with an immedi-ate equity run-up of $10,000 and a subse-quent close out of the trade with a profit of$10,000. You could then say that the tradewas 100 percent efficient. The sign of agood system is the clustering of small los-ing trades with a low run-up in equityand a clear line of winning trades in a 45-d e g ree angle. The closer the trades hugthe 45-degree line, the more efficient thesystem is. However, most systems do notdisplay 100 percent efficient trades andconsequently end up giving back much ofthe open pro f i t s .

Studying a chart like this is an excel-lent way of evaluating the relative effi-ciency of individual trades. For onething, you should be able to detect indi-vidual trades where it would have beenbeneficial to capture profits before thesystem would have closed out thetrades. By doing this, you are maximiz-ing the trade before the momentumfades away.

For greater flexibility, you also couldsplit the analysis between long and shorttrades. (As it stands, you don’t know if aparticular trade was long or short.) Doing


50

40

30

20

10

0

-10

-20

-30

-40

-50

54

32

$1,000

$2,500$2,000

$1,500

$3,000

$4,500$4,000

$5,000

$3,500

The peaks and troughs of these two charts show which profit tar -gets work best and when it is a good idea to add on to a trade.

FIGURE 2 PROFIT TARGET DIFFERENCES

Days in trade

Profit targets

1.60

1.40

1.20

1.00

.80

.60

.40

.20

0

54

32

$1,000

$2,500$2,000

$1,500

$3,000

$4,500$4,000

$5,000

$3,500

Days in trade

Profit targets

A

B

this you also would be able to detect if thesystem is likely to perform better on aparticular side of the market. Forinstance, an MFE chart with most of thetrades clustered in the lower left cornerusually implies the market is not particu-larly prone to tre n d i n g .

In the case of this particular system,there is an abundance of winning tradesc l u s t e red between the $10,000 and$30,000 maximum open profit tradingrange. (There also were three trades witha maximum run-up exceeding $30,000,but we’ll leave them out of the discus-sion for now.) In fact, there are 19 win-ners and five losers within that openequity range. Comparing each trade’smaximum open profit with its endresult, it becomes obvious that trades inthat dollar range give back a large per-centage of their gains by the time thetrade is finally closed out.

N o w, consider what the result wouldbe using a profit target of $10,000. As youcan see, there are five losing trades thatgenerated a total net loss of $40,000. Wi t ha $10,000 profit target, you would haveconverted all of them into winners andthe $40,000 net loss would now be a gainof $50,000. That would add $90,000 toyour account equity and would havereduced the drawdown.

However, at the same time, out of theother 19 trades that turned out to be win-ners, eight trades originally would havemade more than $10,000, for a total of$109,000. If we use a $10,000 profit targeton these trades, they would have onlymade $80,000, thus, decreasing the bot-tom line by $29,000 compared to theoriginal system. For the 11 remainingtrades, which only generated a total netprofit of $66,000, using a profit target of$10,000 per trade would have generateda total net profit of $110,000, an increaseof $44,000.

For all trades the total net pro f i twould have increased by $37,000 to$85,000, with the average profit per tradegoing from $457 to $825. Figure 4 com-pares the equity curve for the originalsystem (in red) and the equity curve forthe system with the profit target added(in blue).

But how do we handle the remainingthree trades on the far right of Figure 3?In statistics, these trades are called “out-liers” because they are more than threes t a n d a rd deviations away from thevalue of the average trade and, as such,

are considered “freak occurrences” andtherefore not used in our analysis.

Nonetheless, there will be times whenyou will miss a really big move usingprofit targets, but overall, using a profittarget should have a soothing effect onyour nerves while at the same timeincreasing your bottom line. In fact, even

though you had to treat those three bigwinners with only a $10,000 profit pertrade, you still would have ended up netpositive, because when the market is in atrending mode a profit target will allowyou to break up a huge winning tradeinto several smaller ones, something thatis not accounted for in this analysis.Ý


Source: Portfolio Maximizer

By comparing each trade’s maximum open profits with its end result you canget a feel for where to take profits before the momentum of the trade turnsagainst you.

FIGURE 3 PROFIT TARGET LIKELIHOOD

Maximum favorable excursionStop and Reverse System (1/1/99-6/29/00)

30,000

25,000

20,000

15,000

10,000

0

Run-up (in $)

0 10,000 20,000 30,000 40,000 50,000

Source: Portfolio Maximizer

Not only will a profit target increase your final equity, it also is very likely todecrease the drawdown, making the strategy easier to trade.

FIGURE 4 EQUITY IMPROVEMENTS

System enhancement with profit targets

Without profit targets

With profit targets

$120,000

$100,000

$80,000

$60,000

$40,000

$20,000

0

-$20,000

-$40,000

-$60,000

-$80,000

W hile momentum (orprice momentum) is ageneric term used todescribe the speed or

force of price movement, it also is thename of a specific calculation that meas-ures price change over a given period.Rate of change (ROC) is simply an alter-nate calculation. The implications andinterpretations of these two studies areidentical.

Momentum and ROC are similar to

oscillators such as the relative strengthindex (RSI) and stochastics in that theyare generally used to highlight shorter-term price momentum extremes (over-bought or oversold points).

The most common calculation for mo-mentum is simply today’s price (typical-ly the closing price on a daily chart)minus the price n days ago (P t o d a y - Pn d a y s

a g o). The most basic rate of change formu-la is today’s price divided by the price ndays ago (Pt o d a y / Pn days ago). Alternate cal-culations for rate of change are1 0 0 * ( Pt o d a y/ Pn days ago) or (Pt o d a y - Pn days ago) / Pn

days ago. The resulting indicators are the

same, though.Figure 1 compares 10-day momentum

and ROC studies and shows that, exceptfor scaling, the two calculations are iden-tical. Momentum simply expresses pricechange as the difference between twoprices while ROC expresses price changeas a percentage or ratio.

Momentum and ROC have “equilibri-um” or center lines that reflect neutralprice momentum (when the curre n tprice is the same as it was n days ago). Inthe case of the momentum study, thisline has a value of zero; for the ROCstudy, the line has a value of 1 (or 100, or0, depending on the calculation). Wheneither of the studies is above the equilib-rium line and rising, price is acceleratingto the upside. When the studies arebelow the equilibrium line and falling,price is accelerating to the downside.

These studies can be calculated onany time frame. Figure 2 shows a six-barmomentum study applied to a 10-minute chart.

O v e r b o u g h t - o v e r s o l d : M o m e n t u m /ROC are typically used to highlight short-e r-term swing points. High momen-tum/ROC readings reflect a potentiallyoverbought market, while low re a d i n g sreflect a potentially oversold market. BothF i g u res 1 and 2 show how many of thehigh and low readings on the indicatorsc o r responded to swing points in the stock.

A c c o rd i n g l y, traders can look forsigns of price reversals when these indi-cators move to extreme levels. However,as will be described in the next section,extreme momentum/ROC readings donot guarantee price reversals or forecastthe degree of the subsequent move; theysimply alert the trader that reversals orcorrections may occur.

Divergence: Another application ofthese studies is to look for re v e r s a l swhen the indicator diverges from price.


Indicator Insight: M o m e n t u mand rate-of-change


Momentum and rate-of-change (ROC) are different calculations of the samestudy. Both measure price change by comparing the most recent price to theprice a certain number of days in the past.

FIGURE 1 MOMENTUM AND ROC

10-day momentum Overbought

Equilibrium (“zero”) line Oversold

Motorola Inc. (MOT), daily

10-day ROC Overbought

60

40

30

50 149⁄256

10.000

0.000

-7.479

4 11 18 25 1 8 15 22 29 6 13 20 27 3 10 18 24 31 7 14 22 28 6 13Oct. Nov. Dec. Jan. 2000 Feb. Mar.

20.000

0.000

-12.881Equilibrium (“zero”) line Oversold

TRADING Basics

For example, in an uptrend price maymake a higher high but the indicatormay make a lower high suggesting thenew high is occurring on weakermomentum and the market is vulnerablefor a reversal. Figure 3 shows how Dell(DELL) made a higher high while theROC study made a lower high. This issometimes called a bearish divergencebecause it suggests the possibility of adownside reversal. In this case, a sell-offdid, in fact, occur.

One of the most important aspects ofindicators such as momentum and ROCis that they measure the acceleration ofprice over time. For example, a marketthat gains a point a day over a givenweek is steadily rising, but its momen-tum, while positive, is actually stagnant.A momentum or ROC study trackingsuch a market would actually be movinghorizontally. By comparison, a stock thatgains a half-point one day, a point thenext, and two points the third day, haspositive and increasing momentum — inother words, the stock is accelerating.This example would be reflected by anupward sloping momentum/ROC line.

As a result, momentum/ROC (andother momentum-based oscillators) willsometimes “lead” the market, reversingbefore price itself. When a market con-tinues to rise, but more slowly (as isoften the case when a price trend isbeginning to exhaust itself), its momen-tum is decreasing, which will be reflect-ed by a down-turning momentum line.As a result, the momentum study maybe falling when price is still rising. Inextended trends, however, this processmay be very drawn out: Price may makea series of higher highs while themomentum/ROC study makes lowerlows (“multiple divergences”) beforefinally reversing in earnest.

The degree to which these studiesreflect longer- or shorter-term priceaction depends on the number of daysused in their calculations. For example, a5-bar ROC will reflect fairly short-termprice swings. A 40-bar ROC will reflectlonger-term price swings. Figure 4 com-pares 5- and 40-bar momentum studies.The 40-bar study tends to follow thelonger-term trend while the 5-bar studyaccentuates more of the shorter- t e r mprice swings.

Unlike oscillators such as the re l a t i v e

s t rength index (RSI) or stochastics, whichrange between set boundaries of 0 and100, momentum/ROC fluctuations arenot limited. As a result, what constitutesan overbought or oversold reading canonly be determined relative to the study’s

past highs and lows. For example, theoverbought and oversold levels in Figure5 were established at levels that capture dmost of the past indicator extre m e s .Overbought and oversold levels can beadjusted to capture m o re or fewer of the



Here, the ROC study moves lower while price moves higher — a bearish divergence that signals potential weakness.

12-day momentum (ROC)

Higher price highs

Lower momentum highs

Dell Computer (DELL), daily

52

48

44

40

42 1⁄8

1 8 15 22 30 5 12 19 26 3 10 17 24 31 7May June July Aug.

20.000

0.000

-20.000

-9.346

42 7⁄16


Momentum/ROC can be applied on any time frame. Here, a six-bar momen -tum study follows the stock’s intraday swings.

FIGURE 2 INTRADAY MOMENTUM

Six-bar momentum

Dell Computer (DELL), 10-minute

43

42 1⁄2

42

41 1⁄2

42 1⁄8

12 13 14 15 10 11 12 13 14 15 10 11 12 13 14 15 10 11 12 138/8 Tuesday 8/9 Wednesday 8/10 Thursday

1.000

0.003

-1.000

i n d i c a t o r’s extremes depending on at r a d e r’s level of activity.

Most traders establish overboughtand oversold levels equal distancesabove and below the equilibrium line,but in many cases this is inappropriate.Overbought and oversold levels willchange over time given the strength andd i rection of the trend. For example,Figure 4 shows how during the down-trend, the longer-term study especiallyis skewed to the downside — it flattensout and is pushed below the equilibri-um line while the shorter-term studystill manages to oscillate above andbelow the equilibrium line. As a result,during uptrends it is generally useful toshift overbought and oversold lineshigher, and shift them lower duringdowntrends. Doing so will cut down onthe number of “false” signals — repeat-ed overbought signals in an uptrend oroversold readings in a downtrend —and increase the number of signals inthe direction of the trend.

Afinal word on trend: Because of theinfluence strong trends exert on indica-tors of this type, momentum/ROC over-bought and oversold signals are moreuseful when the market is moving side-ways in a congestion period or tradingrange. (For example, note the pre m a t u rebullish divergence in April in Figure 5.)Very extreme overbought and oversoldreading accompany strong price thru s t sthat indicate trend strength rather thanexhaustion. Because of the risk of falsesignals, many traders use momen-tum/ROC to alert them to potentialoverbought and oversold situations,then look for confirmation from priceaction to actually enter a trade.

Momentum/ROC are simple studiesthat provide a snapshot of price direc-tion and velocity. They are generallyused to identify overbought and over-sold levels — points at which price islikely to correct or reverse. Divergencebetween price and these indicators canbe used to highlight weakeningmomentum in a trending situation.

These are not systematic tradingtools. Traders should not automaticallysell when one of these studies registersoverbought or buy when they are over-sold. Instead, momentum/ROC canalert traders that a market may be at amomentum extreme.Ý



Five- and 40-day momentum studies reflect the shorter- and longer-termprice swings. Notice that the 40-day study is influenced much more by thelonger-term trend than the 5-day study.

FIGURE 4 SHORT-TERM, LONG-TERM

Five-bar momentum

40-bar momentum

Microsoft (MSFT), 60-minute 84

80

76

72

68

26 3 10 17 24 31 7July Aug.

10.000

0.000

-10.000

5.000

0.000

-5.000

-9.346

72 13⁄16

3.000

2.781


Horizontal lines based on previous indicator highs and lows mark overboughtand oversold levels. Also note the extended bullish divergence that formed inApril and May as the stock dropped while the 10-day rate of change movedhigher.

FIGURE 5 OVERBOUGHT AND OVERSOLD

Overbought10-day ROC

Oversold

Microsoft (MSFT), daily

110

100

90

80

70

28 6 13 20 28 3 10 17 24 1 8 15 22 30 5 12 19 26 3 10 17 24 31 7Mar Apr. May June July Aug.

20.000

0.000

-20.000

72 13⁄16

3.000

4.762

BY MICHAEL A. MERMER

D espite what most peoplethink, the real stock marketcrash that ushered in theGreat Depression occurred

not on Oct. 29, 1929, but between 1930and 1932. The so-called “Crash of 1929”most people refer to, when the marketfell some 50 percent (from 386.10 inSeptember to 195.40 at the end ofNovember), was nothing but a previewof what was to come.

As Figure 1 shows, the real sell-offand bear market did not start until Oct.31, 1930, when the Dow traded belowsupport from the 1929 lows and declinedabout 80 percent, from 195.40 to 40.60 onJuly 8, 1932. Could the recent marketsell-off be a similar preview for a greater

decline to come? There is always thatchance, so familiarizing yourself with afew trading tools and techniques thatcould be handy in such an event is awise move.

Figure 2 shows that, aside from the bigrally from 1924 to the highs in 1929(when the Dow broke through its previ-ous highs at 105.60 and continued torally nearly 300 percent to 386.10), andthe big correction from 1929 to 1932, thetwo most prominent events were the 15-year advance between 1950 and 1965and the current advance that got under-way around 1983. In both instances, theDow increased approximately tenfold.Another significant period was between1965 and 1980 when the Dow stayed in arange from about 600 to 1000, and wasplagued with frequent 30- to 50-percentcorrections.

In late August 2000, the Dow was trad-ing around 11,000, after a rally of nearly1,000 percent since it broke through the1,000 level back in 1983. Given pre v i o u ssituations and based on the market’smassive gains, it’s not unthinkable thatthe current market could experience asimilar correction, especially after factor-ing in that this bull market has now last-ed for close to 18 years without any cor-responding declines.

Yes, the market did “crash” in 1987,when the Dow corrected about 40 perc e n tf rom a high of 2,662.30 on Oct. 2 to the lowat 1,616.20 on Oct. 20. But as Figure 2shows, this correction was not particular-ly dramatic compared to several otherdeclines the market has experienced. Notethat the crash of 1987 only lasted for amonth and that for the entire year of 1988the market never again traded below theOctober 1987 lows.

By contrast, during the 1930 to 1932period the market kept making lowerlows each month for almost two years.During the 15-year consolidation periodt h rough the 1960s and ‘70s there also wasa 50-percent correction, which ended in1974. The bottom line is that the corre c-tion of 1987 was not much of a corre c t i o nby historical standards and nowhere asserious as the 1930 to 1932 dro p .

Based on this historical analysis andthe Dow’s tendency to fluctuate ina p p roximately 15-year intervals, it ispossible that the Dow now could be onthe verge of another extended tradingrange, similar to the one in the ‘60s and‘70s. It has happened before and it canhappen again. But while we can’t be sureit will happen or when, we can be ready.We’ll need to do our analysis homeworkand make sure we know when and whatto trade.

If history is any indication of the future,the current bull market may not contin-ue for much longer. Even if the market ishigher a few years from now, its path isunlikely to be the relatively smooth oneit has been over the last decade.

If the market is indeed likely to chopa round for an extended period, to pre-serve capital, money has to go into bondsand away from the long-only positions inthe typical Internet and high-flyingNasdaq stocks that have made thebiggest gains over the last few years. A n dbecause volatility levels are likely to be


T r a d i n g T H R O U G H T U R M O I LA prolonged period of falling prices or general market turmoil might not be upon us tomorrow, but the odds are that one or the otheris in our future at some point. Read on to find out whatcan you do to prepare.

The Big PICTURE

D

T H R O U G H

high and long and short trade opportuni-ties will be available, you should consid-er trading both sides of the market.

Looking at the last few years’ priceaction in Figure 3a , we can find key sup-port in the Dow at the March 2000 lowand, further down, at the September

1998 lows. Considering the Nasdaqalready has had a 40 percent correctionearlier this year (see Figure 3b, below), asecond sell-off could easily be of the sameor greater magnitude as everyoneattempts to exit at the same time.

During the last bull run many indi-

vidual investors foolishly borro w e dmoney on their credit cards to buystocks. In an effort to rescue their capital,many were forced to sell out at the lowswhen they received margin calls. Thenormal day-to-day volatility of theNasdaq also could accelerate any bigcorrection. If the support lows around3,000 established this May do not hold, itis possible we could go even lower —and there’s still plenty of non-risk capitalinvested in the Nasdaq to fuel a sell-off.

One alternative is to bypass the currentbig downside risk by waiting until themarket has its next big correction, keep-ing money in bonds or short-term treas-ury bills or notes. For now, with CDsgenerating about 7 percent in risk-freereturn, bonds aren’t a bad alternative,considering the current risk level of thestock market. But if we’re in for a 15-yeartrading range, we also need to examine afew alternative strategies.

One of the biggest problems today isthat most traders only see one side of themarket, which has driven the bull mar-ket in part because of a self-fulfillingprophecy — the buyers keep it movingup because it is “supposed” to go up.But what happens when buying stops asinterest rates start to approach 8 to 10percent and inflation sets in? What hap-pens when we start to see earnings fail-ures and it is clear that PE ratios of 50times future earnings are seen as unrea-sonable? That’s when you need to be justas comfortable selling the market as younow are buying it.

Trading in a bear market used to requirespecial knowledge and skills to shortindividual stocks, and for the most partthe practice was limited to institutionaltraders. Also, often individual stockscannot be shorted because they have tofirst be available to borrow and then canonly be sold short on an uptick (see “Awalk on the short side,” Active Trader,July p. 32). This means that if a stock is infree fall without any upticks, you maynever be able to sell it short.

However, with the increasing volatili-ty stemming from a declining market,you might not have to focus on any par-ticular stock, but instead can look at anyof the broader market indices, whichusually provide plenty of pro f i t a b l e


FIGURE 2 THE LAST 80 YEARS

Market rangebound for 15 years

1930 1940 1950 1960 1970 1980 1990

15-year advance

15-year advance

Compared to many other market drops, the crash of 1987 was a relativelyminor event.

1200010000

8000


1000

2000

3000

600050004000

Dow Jones Industrial Index (INDU)-weekly 7/31/2000

FIGURE 1 THE CRASH OF 1930 TO 1932

Dow Jones Industrial Index (INDU)-weekly

1900-1932 crash range

1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934

1929 crash range

The first big drop occurred in October 1929, but the real sell-off and bearmarket did not start until Oct. 31, 1930. The Dow traded below support fromthe 1929 lows and declined about 80 percent, from 195.40 to 40.60, reachingits low on July 8, 1932.

400.00

350.00

300.00


100.00

50.00

150.00

200.00

250.00

trading opportunities in times of tur-moil. There are several ways to tradeindices. One is to trade mutual funds,but this can be costly because of high

fees. A second way is to trade futuresand options, but this requires that youfeel comfortable using leverage and thatyou know how to handle the time decay

and pay the high premiums associatedwith options trading.

A t h i rd and very favorable way are theso-called tracking stocks or index share sthat track three broad market indices —the Dow Jones Industrial Average, the S&P500 and the Nasdaq 100 — and tradet h rough your regular broker or onlinestock account. Their symbols are DIA, forthe Dow, established in 1998; SPY, for theS & P 500, established in 1993; and QQQ, forthe Nasdaq 100 index, trading since 1999.F i g u res 4a through 4c (right), show howclosely these stocks track the actual indices.

With these stocks you’re no longerlimited to the very risky short selling ofindividual stocks or buying expensiveput options. Another key reason to useindex-tracking shares instead of individ-ual stocks is that you can sell them on adowntick.

To trade these stocks efficiently in a trad-ing range or declining market, youshould use limit orders when the marketis slow so you can get your price, andreduce and eliminate the size of anys p read. In a fast market or when thingsbegin to heat up, always use marketo rders, otherwise you’re likely to missthe move. If you’re right on the dire c t i o n ,you will make up the spread or any slip-page on the momentum of the trend any-w a y. Remember, all big winners start outas small ones and some of the very besttrades look lousy when they first begin.Not all winning trades immediately goyour way, so always expect some heat oradverse price movement on a trade.

Also, whether trading in a bull or bearmarket, always trade with a stop. Whilestops are the only way to control risk, notrade is completely without risk.Because markets can gap up or down,stops do not guarantee you’ll get out atyour risk level.

The index-tracking stocks give you theopportunity to use several intere s t i n gtrading strategies, such as spreading onestock against another, in diff e rent quanti-ties, depending on what you believe isthe most likely outcome for the markets.But in order to do so, you first need tofind out how these instruments stack upagainst each other.

The second column in Table 1, showsthat, since Jan. 1, 1990, the Nasdaq index


FIGURE 3 LOOKING FOR SUPPORT

Key support

Key support

The Dow has key support around the 10,000 and 7,500 levels, while supportin the Nasdaq is around the 3,000 level, which already has been tested.

12000

11000

5000

4000

3000

10000

6000

7000

9000

8000

Dow Jones Industrial Index (INDU)-weekly A

B

O 1998 A J O 1999 A J O 2000 A

1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Nasdaq-weekly 4500

4000

1500

1000

2500

2500

3500

3000

Total Av e rage Av e rage Av e ra g e Dull-period return yearly monthly c rash r e t u r n

1990 - 1999 r e t u r n r e t u r n r e t u r nNasdaq 1,557% 32% 2.35% -36% 6.7%DJIA 317% 15.3% 1.2% -26% 6.5%S&P 500 316% 15.3% 1.2% -23% 3.2%

TABLE 1 THE DATA

Basic historical analysis allows you to examine the potential return differencesbetween the different indices and their respective upswings and downswings.



has increased 32 percent per year (orabout 2.35 percent per month), while thec o r responding number for the S&P 5 0 0and DJIA comes out to 15.3 percent (orabout 1.2 percent per month), or appro x-imately half as much as the Nasdaq. Thissuggests that to make the same amountof money in either the S&P 500 or DJIAas in the Nasdaq, you will have to investtwice as much.

H o w e v e r, looking at the fourth col-umn, which shows the average perc e n t-age value of the four largest declines(including the crash of ’87), you can seethat the Nasdaq, on average, dropped 36p e rcent, while S&P 500 and DJIAd ropped 23 percent and 26 perc e n t ,re s p e c t i v e l y. This means that, if youwant to invest in the Nasdaq, but, in caseof a major decline, don’t want to losem o re money than you would have in theother markets, you should only allocatet w o - t h i rds of your available capital.

N o w, say that you have $100,000 in-vested and normally are fully invested inQQQs but would like to hedge against apossible October scare. One way to do itis to liquidate half of your QQQs and usethat money to go short the DIAs. If themarket indeed continues higher at its“normal” rate, then your return forOctober would be $575 ($50,000*0.0235 -$50,000*0.012) — or $8,350 annually,c o m p a red to $2,350 ($100,000 * 0.0235) —or $32,000 annually — being fully invest-ed in Nasdaq. But had the marketcrashed you would have lost only $5,000($50,000*0.26 - $50,000*0.36), compare dto $36,000 ($100,000*0.36) had you beeninvested in QQQs only. Thus, by allow-ing for a decreasing profit potential witha factor of 4 (2,350/575) you haved e c reased your crash sensitivity by a fac-tor of 7.2 (36,000/5,000), assuming thesep roportions stay the same.

The far right column in Table 1 showsthe annual returns for the three indicesduring the relatively dull market periodfrom January 1992 through December1994.

If and when the stock market entersp rolonged period of high volatilityand/or declining prices, you need to beable to operate on the short side just ascomfortably as the long side. The indexshare instruments give both investorsand traders easy-to-use vehicles thatcan hedge risk on longer-term positionsand trade shorter-term fluctuations inthe stock market.Ý

FIGURE 4 LOOK-ALIKES

The index-tracking stocks follow their respective indices very closely, givingtraders the opportunity to trade long- and short-term price moves on bothsides of the market.


11500

11000

10500

10000

9500

9000

8500

8000

7500

115.00

110.00

105.00

100.00

95.00

90.00

85.00

80.00

75.00

Dow Jones Industrial Index,weekly 7/14/2000

DIA, weeklyA

B

C

1999 20001999 2000

1999 2000

1999 20001999 2000

1999 2000

S&P 500 index, weekly 7/14/2000

SPY, weekly

Nasdaq index, weekly 7/14/2000

QQQ, weekly

1550.00

1500.00

1450.00

1400.00

1350.00

1300.00

1250.00

1200.00

1150.00

1100.00

1050.00

1000.00

950.00

4500

4000

3500

3000

2500

2000

1500

120.00110.00100.00

90.00

80.00

70.00

60.00

50.00

40.00

30.00

155.00

150.00

145.00140.00

135.00

130.00

125.00

120.00

115.00

110.00

105.00

100.00

95.00


BY JOHN SALEEBY

o one stock is an island unaffectedby the ocean of other stocks that create

the market. For example, no single pieceof bullish news or chart support line will

bolster a stock during a severely declin-ing market. As a result, it’s imperative toanalyze the trends of diff e rent stockindices to support individual stocktrades.

Market action earlier this year —when the Dow and Nasdaq switchedleadership roles in the market — high-lights the advantages of using an inter-market, macro-analysis approach toi m p rove the probabilities of individualstock trades. Basically, by identifying thesector or stocks responsible for the lead-ership of a particular stock index, youcan find the best trade opportunities at agiven time.

There are two basic elements to thisapproach:

1) using intermarket analysis (com-paring stock indices) to identify the lead-ing stock index at a given time; and

2) identifying and trading the stocksor sector(s) responsible for the strengthof the index.

As long as the stock, sector and corre-sponding index remain in their re s p e c t i v ebull trends, at least a small long positionin each stock is maintained. A d d i t i o n a ltrades can be made when such stocksretrace to support levels, with pro f i t staken at projected resistance points.

As each of the leadership stocksbegins to correct, the leading sector(s)becomes suspect. Unless new sectors inthe same index emerge to lead the mar-ket, the index also becomes suspect.When the leadership sectors and indicesturn bearish, long positions are liquidat-


Oct. Jan. Apr. Jul. Oct. Jan. Apr. Jul. Jan. Apr.1998 1999 2000

FIGURE 1 FOLLOW THE LEADER

Dow Jones Industrial Average (black), Nasdaq 100 (red), daily

DJIA leads market higher

Intermarket support line

DJIA leads market higher again

DJIA upsidebreakout to

close gapNDX begins ascent

to leadership

NDX becomes new leader

NDX breaks support to close gap

Overlaying the DJIA and Nasdaq (NDX) highlights the relative movement ofthe two indices.

Source: CQG for WindowsOct. 28, 1999

Profiting from INTERMARKET ANAL Y S I SThis spring’s Nasdaq sell-off blind-sided many traders.

H e r e ’s how one trader used a top-down intermarket analysis approach

to time the switch from Nasdaq to Dow issues.

NN

The Big PICTURE

ed and short trades are favored.We’ll use actual trades from earlier

this year to illustrate how the interactionof the Dow Jones Industrial Av e r a g e(DJIA) and Nasdaq 100 indices helpedidentify high-probability individualstock trades.

Figure 1 (opposite page) is a macro pic-ture of the market. It is created by settingthe two indices, the DJIA and Nasdaq100, at the same starting point eachJanuary. The chart is not lognormal — apercentage move in one index is notequal to a percentage move in the otherindex. Nor is the chart to scale — a one-point move in the DJIAis not equal to aone-point move in the Nasdaq 100.Instead, congestion areas are overlappedat the beginning and end of the periodbeing analyzed to compare the move-ment of the indices. Overlapping thecongestion of the relevant time framedemonstrates the indices’ movement rel-ative to their performance over the pre-vious time period and one another.

Lognormalizing or scaling the chartdoes not reveal additional informationbecause the DJIA and Nasdaq 100 arecomputed very differently. The Nasdaq100 is a modified capitalization-weight-ed index of the stocks listed on theNasdaq 100. It is designed to limit dom-ination of the index by a few large stockswhile generally retaining the capitaliza-tion ranking of companies. The DJIAis aprice-weighted index for which a one-point move in any component equates toapproximately a five-point move in theindex re g a rdless of the perc e n t a g emovement of the stock or its market cap-italization. Because of the diff e re n tmethods used to calculate the indices,lognormalizing or scaling the chartwould neither reveal relative moneyflows nor relative price performance ofthe component companies.

This chart is designed only to reveal re l-ative price performance of the twoindices. By overlapping the congestion of

January 1998 and 1999, the relative pricea p p reciation for each index is set for theduration of the chart. For example, if theD J I A a p p reciated approximately 1,275points in 1998 and the Nasdaq appre c i a t-ed approximately 850 points in 1998, thischart fixes the ratio of appreciation as 1.5.Thus, each one-point move in the Nasdaqwill be charted as an appreciation of 1.5,while the DJIA remains one point.Because the two indices begin at the samepoint on the chart, the appreciation of thetwo indices relative to their performanceover the previous year as well as eachother can be easily visualized. Thismethod can be used for any time frame.

Figure 1 shows how the DJIA led theNasdaq in each bull move between early1998 and late 1999. After the breakout inJanuary 1998, the Dow outperformed theNasdaq until August of that year.

When a sector or index leads the over-all market to new highs, the disparitybetween the two can become too gre a tand, as a result, the lagging sector orindex will periodically outperform (close

the gap) until valuations return to their“ relative” levels. For example, from Julyt h rough October 1998, the Dow re t r a c e dto the level of the Nasdaq 100 beforereasserting its leadership and taking bothmarkets higher for the next 12 months.The Dow’s leadership role is obvious andconsistent in 1998 and most of 1999.

H o w e v e r, between October 1999 andJanuary 2000, the Nasdaq stopped laggingthe Dow. In fact, by December 1999 theNasdaq had assumed the leadership ro l ein the market, continuing to the upsidewhile the Dow peaked and then declined.

This leadership role reversal was con-firmed by mutual fund flows. Until theend of 1999, flows heavily favored theS&P 500. But by early 2000, S&P indexfunds actually underwent redemptionsas money transferred to technology andgrowth funds.

The Nasdaq’s bullish breakout in lateOctober 1999 is also shown by the


Apr. Jul. Oct. Jan. Apr. Jul. Oct,

1999 2000

FIGURE 2 NASDAQ—DOW SPREAD

Nasdaq 100—Dow spread, daily

Nasdaq 100 breakout

Double top

Confirmation of double top (breaks second trendline),

sell short

Breaks first trendline,sell all Nasdaq positions

The relationship of the Nasdaq 100 and the Dow shown as a spread. Note the double-top formation.

Source: CQG for Windows

M a r. 10, 2000Oct. 28, 1999

To profit from index breakouts, you must determinewhich stocks within the leading index account for the bullish trend.

Nasdaq-Dow spread in Figure 2. Equallyobvious is the reversal of this trend withthe formation of a double top in mid-March 2000. The question is how to usethis information to your advantage

when trading individual stocks.To profit from index breakouts like the

one in the Nasdaq, you must determinewhich stocks within the leading indexaccount for the bullish trend. There are

several ways to do this, including exam-ining the new highs and perc e n t a g egainers on bullish days. (Also, “headlinestocks” — stocks that receive an inordi-nate amount of media and analyst cover-age for a period of time — often becomeshort-term leaders.)

When multiple stocks like these are inthe same sector, you can chart either theindividual stocks or the sector itselfagainst the index. If the sector or indi-vidual stocks ascend at a greater ratethan the index, they are, by definition,leading the market and are good candi-dates for individual trades. (This isessentially a way of graphically target-ing stocks with high relative strength[RS] readings.)

For example, in October 1999 the fol-lowing stocks from the B2B e-commercesector met the criteria mentioned above:Commerce One (CMRC), Ariba (ARBA),Vitria (VITR) and Akamai (AKAM).Each of these stocks (as well as the B2Bsector as a whole) was featured onCNBC almost daily, and they were oftenleading point and percentage gainers.The trendlines (regression lines can alsobe used) for the individual stocks andthe B2B sector were ascending at agreater rate than that of the Nasdaq 100.

In December 1999, the genomic stocks(those biotech companies focusing ongene-based medical treatments) joinedthe B2B stocks in their leadership role.Figure 3 shows the late December break-out in the biotech index (BTK); Figure 4shows the coinciding breakout forHuman Genome Sciences (HGSI), one ofthe leading genomic stocks.

Long positions were initiated at thistime in both HGSI and Celera (CRA),another leader in the sector. These posi-tions were maintained, at least in part,until March 2000. Note that on March 2,HGSI turned bearish (as did CRA onMarch 3), and by March 7, the BTK hadclearly turned down as well.C o n s e q u e n t l y, the positions in thesestocks were liquidated.

By March 10 each of the selected lead-ing equities (except VITR) in the B2B e - c o m m e rce sector and biotechnologysectors had turned bearish, and theabsence of new sectors asserting leader-ship cast a pall over the Nasdaq as awhole. (Also, the Nasdaq 100 and Nasdaq


Nov. Dec. Jan. 2000 Feb. Mar. Apr.25 1 8 15 22 29 6 13 20 27 3 10 18 24 31 7 14 22 28 6 13 20 27 3 10

FIGURE 3 BTK BREAKOUT

Biotechnology Index (BTK), daily

Breakout

Penetration of short- and long-term trendlines

A mid-December 1999 breakout in the biotech (BTK) index coincided with the Nasdaq assuming the leadership role in the overall market.


800.00

700.00

600.00

500.00

400.00

300.00

5 0 5 . 1 8

Jan. 2000 Feb. Mar.13 20 27 3 10 18 24 31 7 14 2 28 6

FIGURE 4 SECTOR LEADER: HUMAN GENOME SCIENCES

Human Genome Sciences, Inc. (HGSI), daily

Breakout

Penetration ofup trendline

Double top

Genomic/biotech leader HGSI was a strong stock in a strong sector, making ita viable candidate for a long position when the Nasdaq and BTK broke out tothe upside in December 1999.


240

220

200

180

160

140

120

100

80

60

1 72 1⁄2

Composite both closed sharply lowerafter establishing new highs, re i n f o rc i n gthe conclusion that at least a correction inthe Nasdaq was imminent.) Conse-q u e n t l y, all remaining long Nasdaq posi-tions were liquidated at this time, antici-pating at least a correction in the Nasdaq.

At this point, only short positions inthe Nasdaq were considered. Given thatthe Nasdaq 100, leadership sectors andeach leadership stock (except VITR, seeFigure 5) had turned bearish, the oddswere good that the expiration of VITR’slockup period would cause that stock tosell off as well. VITR was artificiallyinflated by bullish information that cre-ated retail demand at the end of thestock’s lockup period on March 15 (seeActive Trader, May,). Consequently, uponthe March 13 penetration of the uptrendline shown in Figure 5, a short posi-tion of VITR was established. (The posi-tion was liquidated on March 22 uponupside penetration of the descendingtrendline.)

While the Nasdaq showed signs of

topping in March, the Dow (Figure 1)and S&P 500 (not shown) were bottom-ing. The Nasdaq-Dow spread (Figure 2)broke an up trendline, confirming the

reversal and indicating the two indiceswould begin to close their gap.

But there was more to this than a sim-ple change of leadership. Because theDow and S&P charts indicated bullishbreakouts despite a Nasdaq correction,the odds were good the Dow breakoutwould be explosive. Why? Because tech-nology stocks, mostly listed in theNasdaq, comprise approximately 25 per-cent of the S&P. For the S&P to bullishlyb reak out despite a bearish Nasdaqrequires double the work from the Dow.

As a result, Dow stocks became the log-ical focus of long-side trades, in anticipa-tion of an upside breakout. (Alternately,the Diamonds [DIA] and Spiders [SPY]could be used to take advantage of movesin their respective indices.)

H o w e v e r, positions in Micro s o f t(MSFT), Intel (INTC), IBM (IBM) andHewlett-Packard (HWP) were not goodcandidates because of their correlation tothe Nasdaq. Instead, General Electric(GE), Coke (KO), and American Express(AXP) were excellent choices to partici-pate in the Dow rally: GE has a 99 per-cent correlation to the S&P; KO was nearits 52-week low and undervalued; AXPis a financial leader that had correctedseverely during the Dow downturn.

Mid-morning on March 15, after theDow clearly failed to penetrate the bot-tom of its range (Figure 6), these stocksw e re purchased at the market. March 15


Feb. Mar. Apr.10 18 24 31 7 14 22 28 6 13 20 30 3 10

FIGURE 5 SECTOR LEADER: VITRIA TECHNOLOGY

Vitria Technology, inc. (VITR), daily

Profits takenwhen stock

penetrates downtrendline

Sell-off coincides with end

of lockup period

VITR was a leader in the B2B e-commerce sector that, along with the BTK,was propelling the Nasdaq to new highs.


100

90

80

70

60

50

40

3028 3⁄8

Jan. 2000 Feb. Mar. Apr.27 3 10 18 24 31 7 14 22 28 6 13 20 27 3

FIGURE 6 DOW BOTTOM

Dow Jones Industrial Average (DJIA), daily

Dow fails to penetrate

bottom of trading rangeand insteadbreaks out

to the upside

Biggest back-to-back pointgains in Dow history

Short-term trendline penetrated

The March top in the Nasdaq was reflected by the bottoming pattern in theDow. The Industrials failed to penetrate the trading range’s support and shotto the upside, signaling the end of the Nasdaq’s leadership and a switch toDow stock that were not correlated to the Nasdaq.


11800.00

11600.00

11400.00

11200.00

11000.00

10800.00

10600.00

10400.00

10200.00

10000.00

9800.00

1 1 1 1 1 . 4 8

and 16 were the largest back-to-backpoint gains in Dow and S&P h i s t o r y.

In such situations, if the reversal ofsuch an entrenched trend doesn’t mani-fest itself in a buying frenzy, the rally issuspect. But if the intraday reversal andbuying spree do materialize, the nextday’s follow-through should be equallypowerful. If momentum wanes towardthe end of the second day, positions canbe exited. Trailing stops can be used tolock in profits and allow the position to

run if longer-term follow-through devel-ops. (In this case, the long positions wereexited on a penetration of the short-termup trendline on March 28.)

At some point the previously leadingindex should bottom and attempt toreassert its leadership. After the doubletop and downward penetration of the upt rendline in the Nasdaq-Dow spread on

M a rch 29 (Figure 2), the Nasdaqremained in a downtrend until April 24.As a result, only short positions were con-s i d e red during this period. When theNasdaq sell-off began, the initial pro j e c t-ed low was 3,575, based upon pre v i o u sNasdaq congestion and projected inter-market support (see Figure 1). Notice theextension of the Dow trendline from itsleadership period. This line re p re s e n t sthe extension of Dow valuations over thissame period of time and there f o re sup-

ports the valuation of the leading indexupon a retracement to the tre n d l i n e .

This projection was modified afterApril 4, when the Nasdaq 100 both plum-meted and rallied approximately 500points in a day, and bounced off the inter-market support line. The projected lowwas set with a trendline incorporatingthis plunge. The sell-off on April 14 hit

the trendline support at 3,200 (see Figure7). On April 17, after a 20 percent re v e r s a lof the intraday low set at the open, longpositions in both the QQQs (the Nasdaq100 tracking stock) and a pre d e t e r m i n e dbasket of stocks were executed. April 17and 18 were the largest back-to-backgains in Nasdaq history.

A more thorough application of this kindof analysis would encompass other relat-ed indices — e.g., following the trends inthe treasury markets and their positiverelationship with the equity markets.Simply, what is good for the treasurymarket is generally good for the stockmarket; movement in treasuries oftenprecedes movements in stocks.

Also, when examining the tre a s u r ymarket, you should consider the trend ofboth the market and the yield curve todetermine the effect on equities. For exam-ple, while a flattening yield curve is gener-ally bearish for equities, the true eff e c tdepends on the cause of the flattening.

Most yield curve flattenings are per-ceived to be bearish because short-termi n t e rest rates are rising. However, if thecurve is flattening because the longer- t e r myields are dropping, this can be bullish forequities, especially when accompanied bya downward shift in the yield curve.

Equally important are the currencymarkets (recall the Thai baht crisis, in thefall of 1997), commodity indices (espe-cially gold and oil) and foreign equitymarkets. Understanding the direction ofeach of these markets and their effect oneach other will greatly enhance yourability to forecast the direction of theU.S. stock market and consequently,individual stock trades.

R e m e m b e r, no stock is an island.Trading individual stocks without exam-ining broader market trends is likefocusing on an individual raindrop andignoring the downpour.

For further reading: “Intermarket Technical Ana -lysis,” John J. Murphy. (1991, John Wiley & Sons,New Yo r k . )

Ý


T r a d i n g individual stocks without examining broader market t r e n d sis like f o c u s i n g on an individual raindrop and ignoring the downpour.

Aug. Sep. Oct. Nov. Dec. Jan.2000 Feb. Mar. Apr. May Jun.

FIGURE 7 NASDAQ BOUNCE

Dow Jones Industrial Average (DJIA), daily (black)

Nasdaq 100 Index (NDX), daily (red)

DJIA-NDX intermarket support line

NDX breaks supportto close gap

NDX bouncesoff support

DJIA upside breakoutto close gap

The Nasdaq 100 plunged, bounced quickly, then dropped again before rallyingoff the trendline support (that incorporated the plunge) around 3,200.


Oct. 28, 1999

Markets: Stocks, stock index futures andindex shares (SPDRs, DIAs, QQQs)

System logic: This is a short-term relative-s t rength (RS) system that compares thestock price with its underlying marketindex, but also looks for confirmation fromvolume.

The first step is to calculate the relativestrength of the stock compared to the mar-ket as a whole, then take the percentagedifference between the relative strengthcurve and its five-day moving average. Avalue above one (the higher the better) indicates the stock hasmoved stronger than the market as a whole.

The second step is to calculate the percentage differencebetween the on-balance-volume indicator and its five-daymoving average. A value above one (the higher the better)indicates that the volume fueling the move is above its five-day moving average and that the market has a short-terminterest in the stock.

The final step is to multiply both the percentage differencesto create a volume-weighted relative-strength curve. A valueabove one (the higher the better) indicates that the marketboth has a short-term interest in the stock, but also that it isdoing better than average in terms of the overall market. (Or,if the interest is low, the stock is still moving strongly enoughto warrant a trade and vice versa.)

Rules:1. Enter long at the close when the volume-weighted relative-strength curve crosses above 1.

2. Risk 0.5 percent of available equity per trade. At the time ofe n t r y, calculate the number of shares per trade as 0.005 multi-plied by available equity, divided by the dollar value of the dis-tance between the entry price and a 4 percent stop-loss.

3. a. Exit at the close when the volume-weighted relative-strength curve crosses below 1.

b. Exit with a loss if the market falls below the entry priceminus 4 percent.

Test period: Dec. 11, 1985, to July 12, 2000

Test data: Daily stock prices for the 30 mostactively traded Nasdaq 100 stocks, excludingthose stocks that also can be found in otherindices, such as Microsoft and Intel in the DowJones Industrial Average and S&P 500 index.$20 per trade has been deducted for commis-sion.

Starting equity: $100,000 (nominal)

System drawbacks:This is a long-only system, which makes it vul-nerable to market corrections. The systemprobably would benefit from the addition of atrend filter (such as a long-term moving aver-age) that would only allow trades in the direc-tion of the prevailing trend. Developing a simi-lar strategy for the short side would make itpossible to make money no matter what direc-tion the market is heading.

Other observations:Although this system does better than buy-and-hold for both the Dow and S&P 500, at


The TRADING Systems Lab

$10,000,000

$1,000,000

$100,00012/11/85 12/11/87 12/12/89 12/12/91 12/13/93 12/13/95 12/12/97 12/14/99

EQUITY CURVE

RS system No. 1

SAMPLE TRADES

75

70

65

60

55

50

45

40

April May June July

Apple (AAPL), daily

Sell 1

Buy 1

Source: Omega Research

first glance it doesn’t quite seem to keep up with the Nasdaq100. However, looking closer at the numbers reveals that therisk-reward ratio (total return divided by maximum draw-down) for a buy-and-hold strategy for the Nasdaq 100 comesout to 68.6. For this system, only risking 0.5 percent of avail-able equity per trade, the same number comes out to 80.8,which is a clear indication that it allows you to make moremoney while risking less.

Another favorable aspect of the system is that a buy-and-hold strategy is always in the market, risking 100 percent ofavailable equity. While being in the market all of the timemight not add to the risk, it sure adds to the agony. Lastly,some of the 30 stocks in the test portfolio were not listed until

very recently (such as Amazon.com, whichdidn’t start contributing to the portfolio equityuntil August 1997).

Buy-and-hold stats: DJIA: Total return —- 636 percent;

Max DD — 41 percent; Longest flat — 24 months

S&P 500: Total return —- 640 percent; Max DD — 36 percent; Longest flat — 23 months

Nasdaq: Total return — 2,814 percent; Max DD —41 percent; Longest flat — 19 months


Disclaimer: The Trading System Lab is intended for educational purposes only to provide a perspective on different market concepts. It is not meant to recommend orpromote any trading system or approach. Traders are advised to do their own research and testing to determine the validity of a trading idea. Past performance does notguarantee future results; historical testing may not reflect a system’s behavior in real-time trading.

DRAWDOWN CURVE

0.00%

-5.00%

-10.00%

-15.00%

-20.00%

-25.00%

-30.00%

-35.00%

-40.00%

12/11/85 12/11/87 12/12/89 12/12/91 12/13/93 12/13/95 12/12/97 12/14/99

Source: CSI, Unfair Advantage

Profitability Trade statisticsEnd equity ($): 2,825,018 No. trades: 9,444Total return (%): 2,725 Avg. trade ($): 2,272Profit factor: 1.29 Largest loss ($): 47,488Avg. tied cap (%): 53.59 Win. trades(%): 37.58Win. months (%): 63.64 Avg. days: 3.69

Drawdown TIM (%): 98.29*/30.46**Max DD (%): 33.71 Tr./Mark./Year: 21.71Longest flat (m): 11.54 Tr./Month: 54.28

*For the entire portfolio. **Average per market.

STRATEGY SUMMARY

LEGEND: End equity ($): equity at the end of test period • Total return (%):total percentage return over test period • Profit factor: gross profit/gross loss •Avg. tied cap (%): average percent of total capital tied up in open positions •Win. months (%): percentage profitable months over test period • Max DD (%):maximum drop in equity • Longest flat: longest period, in months, spentbetween two equity highs • No. trades: number of trades • Avg. trade ($):amount won or lost by the average trade • Largest loss ($): largest losing trade• Win. trades (%): percent winning trades •Avg. days: average trade length •TIM (%): amount of time there is at least one open position for entire portfolio, onaverage per market, respectively • Tr./Mark./Year: trades per market per year •Tr./Month: trades per month for all markets

Send Active Trader your systemsIf you have a trading system or idea you’d like to see tested inthe Trading System Lab, send it to us. We’ll test it on a portfolioof stocks or futures (for now, maximum 30 markets, using dailydata starting January 1, 1980), using true portfolio analysis/opti-m i z a t i o n .

Most system testing software only allows you test one marketat a time. Our system testing technique lets all markets sharethe same account and is based on the interaction within theentire portfolio.

E-mail your system logic and a short description to [email protected], and we’ll get back to you.

ROLLING TIME WINDOW RETURN ANALYSIS

Cumulative 12 24 36 48 60 months months months months months

Most recent: 17.24% 126.21% 259.10% 507.49% 542.78%

Average: 30.34% 74.02% 128.93% 190.40% 280.00%

Best: 135.97% 364.36% 661.03% 746.72% 1,034.50%

Worst: -6.97% -2.25% 1.96% 15.55% 24.90%

St. dev: 30.95% 76.29% 135.56% 183.82% 275.33%

Annualized 12 24 36 48 60 months months months months months

Most recent: 17.24% 50.40% 53.13% 56.99% 45.08%

Average: 30.34% 31.92% 31.80% 30.54% 30.60%

Best: 135.97% 115.49% 96.70% 70.58% 62.54%

Worst: -6.97% -1.13% 0.65% 3.68% 4.55%

St. dev: 30.95% 32.78% 33.06% 29.80% 30.28%

LEGEND: Cumulative returns — Most recent: most recent return from start toend of the respective periods • Average: the average of all cumulative returnsfrom start to end of the respective periods • Best: the best of all cumulativereturns from start to end of the respective periods • Worst: the worst of all cumu -lative returns from start to end of the respective periods • St. dev: the standarddeviation of all cumulative returns from start to end of the respective periods

Annualized returns — The ending equity as a result of the cumulative returns,raised by 1/n, where n is the respective period in number of years

october 2000 issue 2 3 contributors 4 6 ... - forex-warez.comforex-warez.com/trading books/magazine...

Documents