At January 2008 Issue

PUTTING A PIVOT-DAY SYSTEM TO THE TEST p.
22
INSIDE INFORMATION
STEPHEN MCCLELLAN PULLS BACK
the Wall Street curtain p. 47
TRACKING INSIDER BUYING p. 14
SEA CHANGE FOR T-NOTES? REVERSAL OF FORTUNE:
Find out if the trend Symmetrical reversal
has turned p. 34 days p. 30
TRADING THE EQUITY CURVE

with a twist p. 38
$4.95 U.S. / CANADA Printed in the U.S.A. www.activetradermag.com

CONTENTS:
JANUARY 2008 • VOLUME 9, NO. 1
®
TRADING STRATEGIES FOR THE FINANCIAL MARKETS
Market Pulse
8 The inside track
Find out how Dow stocks have performed after company
“insiders” took positions in them.
By David Bukey
Trading Strategies
13 Designing and testing a pattern-based trading system
This system identifies potential reversals around pivot highs
and lows.
By Howard Bandy
19 Symmetrical reversal bars

When price closely mimics the behavior of the previous day
— in reverse — is it a sign of an imminent price move?
By Active Trader staff
Advanced Strategies
22 Bonds and the first rule of trading
Where do we stand after a 25-year bond bull market? Get
ready to adjust your T-bond and T-note strategies.
By Howard L. Simons
26 Trading System Lab

Reverse-trade equity management
How will trading a system based on the direction of
its equity curve affect performance?
By Volker Knapp
On the Market
30 Why is the sky blue … and why do traders trade?
Do traders lose money because they are overconfident,
or because they pick the wrong stocks? It might be both.
By Active Trader staff
In every issue…
39 Global Marketplace 60 Trading Resources
International market performance. New products, services, and
books.
3 Editor’s Note 40 ETF Snapshot
4 Contributors
Volume, volatility, and momentum 62 Trading Calendar
statistics for exchange-traded funds.
5 Letters 64 Key Concepts
41 Futures & Stocks
6 Opening Trades Snapshot 65 Upcoming Events
Volume, volatility, and momentum
statistics for futures and stocks.
1 www.activetradermag.com • January 2008 • ACTIVE TRADER

Contents continued
Contact Active Trader:

Editorial inquiries: editorial@activetradermag.com
Comments, suggestions:
suggestions@activetradermag.com
For advertising or subscription
information, log on to: www.activetradermag.com
© 2007, Ethan Pines
Active Trader Interview

33 Behind the Wall with Stephen McClellan
A veteran stock analyst spills the beans on
what goes on behind the scenes on Wall
Street — and how individual investors get The Economy
steamrolled by misleading information. 52 U.S. Economic Briefing
By Mark Etzkorn Updates on economic numbers and how
the market reacted to them.
38 The Face of Trading
Lifelong trader Technology for Traders
By Active Trader staff 54 Software screening: MultiCharts 2.1
Reviewed by David Bukey
Inside the Market
By Jeff Ponczak Business of Trading
58 A trader tax primer
42 Hedge-fund group makes Here’s a refresher course on making the best
recommendations of your tax situation.
Facing pressure from the SEC and By Robert A. Green, CPA
Congress, the hedge-fund industry took
matters into its own hands and created a Trade Diaries
set of guidelines for managers. 66 Whipsawed! Volatile market takes its toll.
Other stories: 67 Short hedge results in worst of both worlds.

Morgan Stanley faces lawsuits • Quick
scalps • Nasdaq, Philly agree to deal • BATS
wants to spread wings • Lawsuits continue
at Refco • Managed money • Trading
restrictions ended at NYSE • NYSE, BIDS
hook up • Global news

Editor’s NOTE
Bubble, bubble, toil and trouble
F rom Jan. 1, 2000 through Oct.

24, 2007, the Shanghai
Composite Index (SSEC)
quadrupled in value on a closing basis;
the S&P 500 index gained 3.18 percent
A stock market that
can rally 70 percent
in a little more than
belt. This stock-trading stuff is still pretty
new to them.
On Thursday, Oct. 25 the world found
out China had enjoyed another quarter of
double-digit growth (11.5 percent).
during the same period. There’s nothing worse than nay saying a
From the Aug. 16 global equity market two months has huge stock rally — shouting “the sky is
low through Oct. 26, the iShares falling” for months (or years) on end as
FTSE/Xinhua China 25 Fund (FXI), an
the potential for the market moves inexorably higher, and
exchange-traded fund (ETF) that tracks ultimately claiming you were correct
the performance of the 25 largest Chinese when a big correction or turnaround
stocks traded on the Hong Kong Stock
proportionally large finally does happen.
Exchange, gained 70 percent — nearly Since the spring, at least, there have
10 times the S&P 500’s 7.4-percent gain. downdrafts. been numerous warnings, including one
The FXI was averaging more than four from Alan Greenspan, about the potential
million shares a day in volume at the end century, I have no idea. (I have owned for a crash in the Chinese market.
of October. both of the previously mentioned ETFs, They’ve all been premature.
But how much do you know about the although I did not at the time of this If China really is destined to be the
stocks in this fund? Do you know what writing.) But I do know when people power of the 21st century, the Shanghai
sectors they represent? Do you know any- flock to something without any real index will someday be much higher than
thing about how the Chinese markets understanding of it, be it technology it is now. (It was only a little more than
work — what regulation they have, or stocks, real estate, or the restaurant busi- 20 years ago that the Dow Industrials tra-
don’t have? Do you know the difference ness, bad things can happen. versed 1,000 for the first time.) But a
between this fund and the Powershares In this case, Chinese individual stock market that can rally 70 percent in
Golden Dragon Halter USX China Fund investors are the ones doing some of the a little more than two months has the
(PGJ)? Did you know that foreign most intense flocking, at least in the case potential for proportionally large down-
investors are prohibited from trading the of the Shanghai index. One can only drafts. Investors and traders have to
stocks in the Shanghai Composite index? speculate about the level of investor know whether they can stomach the
Let’s be honest: The only thing most of sophistication and the amount of emo- roller coaster ride.
us know is that China is hot. No doubt tional speculation involved. (And we who
about it, it has been the place to be for at live in glass houses should not be hurling
least the past two years. Some day it will stones.) But it’s not as if the country has a
not be. Whether that day comes next long history of financial markets and a
month, next year, next decade, or next generation of seasoned investors under its Mark Etzkorn, Editor-in-chief

This Month’s
® CONTRIBUTORS
Dr. Howard Bandy has degrees in mathematics, physics, engineering, and

computer science. He was a university professor of computer science and mathe-
For all subscriber services:
Active Trader Magazine matics, vice president and designer of the major product for a company that pro-
P.O. Box 567
Mt. Morris, IL 61054-0567 duced programs for stock selection and timing, and senior research analyst for a
• commodity trading advisor, where he held a Series 3 license. He is the author of
(800) 341-9384
• the book, Quantitative Trading Systems, which expands on topics discussed in this
www.activetradermag.com
month’s article. Information about the book is available at www.quantitativetrad-
Editor-in-chief: ingsystems.com. Dr. Bandy can be contacted at howard@quantitativetradingsys-
Mark Etzkorn
metzkorn@activetradermag.com tems.com.
Managing editor:
Molly Flynn
mflynn@activetradermag.com
Howard L. Simons is president of Rosewood Trading Inc. and
Senior editor:
Jeff Ponczak a strategist for Bianco Research. He writes and speaks frequently on
jponczak@activetradermag.com
a wide range of economic and financial market issues.
Senior editor:
David Bukey
dbukey@activetradermag.com
Associate editor: Robert A. Green, CPA, is CEO of Green & Company CPAs,
Chris Peters
cpeters@activetradermag.com LLC (GreenTraderTax.com), a CPA firm focused on traders and
Contributing writers: investment management businesses. He is the author of The Tax
Thom Hartle, Howard L. Simons, Marc Chandler, Guide for Traders (McGraw-Hill, 2004). GreenTrader provides tax
Keith Schap, Thomas Stridsman, Robert A. Green,
Jim Kharouf preparation, accounting, consulting, entity, and retirement-plan
Editorial assistant and Webmaster: formation services; IRS/state tax exam representation; and sells trade-accounting
Kesha Green
software. GreenTrader also provides a full line of services to hedge funds includ-
Art director:
Laura Coyle ing launch services, compliance, accounting, software, and tax. For more informa-
lcoyle@activetradermag.com tion or to participate in free conference calls, chat rooms, and message boards,
President: visit www.greencompany.com or call (877) 662-2014 or (646) 216-8061.
Phil Dorman
pdorman@activetradermag.com
Publisher,
Ad sales East Coast and Midwest: Volker Knapp has been a trader, system developer, and
Bob Dorman
bdorman@activetradermag.com researcher for more than 20 years. His diverse background encom-
Ad sales passes positions such as German National Hockey team player,
West Coast and Southwest only:
Allison Ellis
coach of the Malaysian National Hockey team, and president of
aellis@activetradermag.com VTAD (the German branch of the International Federation of
Classified ad sales: Technical Analysts). In 2001 he became a partner in Wealth-Lab Inc.
Mark Seger
seger@activetradermag.com (www.wealth-lab.com), which he is still running.
Volume 9, Issue 1 Active Trader is published monthly

by TechInfo, Inc., 161 N. Clark Street, Suite 4915,
Chicago, IL 60601. Copyright © 2008 TechInfo, Inc. All
rights reserved. Information in this publication may not Jim Kharouf is a business writer and editor with more than 10 years of experi-
be stored or reproduced in any form without written
permission from the publisher. Annual subscription rate ence covering stocks, futures, and options worldwide. He has written extensively
is $59.40.
The information in Active Trader magazine is intended

on equities, indices, commodities, currencies, and bonds in the U.S., Europe, and
for educational purposes only. It is not meant to rec- Asia. Kharouf has covered international derivatives exchanges, money managers,
ommend, promote or in any way imply the effective-
ness of any trading system, strategy or approach.
Traders are advised to do their own research and test-
and traders for a variety of publications.
ing to determine the validity of a trading idea. Trading
and investing carry a high level of risk. Past perform-
ance does not guarantee future results.

LETTERS
Straight answer
on linear regression
In “Short-term T-Bond Trading”
(Active Trader, October 2002), you
refer to a slope line derived from a
linear regression calculation. I use
TradeStation and don’t see it listed
as an analysis technique, indicator,
etc. Is it under another name, or
What a coincidence does it have to be written in
EasyLanguage? Or, can I substitute
The Trade Diaries in the October issue of Active Trader (p. 87 and 88) outlined a short-
another indicator that will give me
ing strategy and what stunned me was the similarity with what I am doing. I have read
the same input view?
a zillion books, been to a zillion shows, and heard a zillion strategies, but the articles
[described an approach] that is identical to what I am doing in my private portfolio.
–Randy G.
For example, prior to market close yesterday, I saw that BBT was faltering on a
strong market. I shorted at 41.96 and as I write, it is down 2.09 percent. I keep doing
TradeStation has two linear regres-
this all the time (I’m serious).
sion tools in its indicator section:
I gave up a CEO position in June to trade a small portfolio. My CEO status did not
“linear reg line” and “linear reg
pay me the sums received by John Chambers and Larry Ellison, but I was confident
curve.” You can access them by
enough to risk a “secure” job for the rewards and risks of the market.
selecting Insert Analysis Techniques
In the last week I saw the exponential rises of SHLD and GRMN, and I consistently
> Indicators.
keep seeing these moves. I appreciate that the market can humble traders, but I have
increased confidence in my abilities, regardless of market direction. (Yes, I am perfectly
comfortable shorting and, in fact, find it much easier to predict.)
Who wrote the Trade Diaries? I would like to approach him/her to see if I can trade Finding articles
alongside them for a day or two. I would love to find a trader who is successful, and
who shares my strategies. I have looked around for months, and though I find many I am a little frustrated finding an index to
wheelers and dealers, the article took me aback, in that it emulated what I do. I do, past reports. Specifically, I am looking for
however, need tutelage. articles on placing spread orders.
Is there a publication containing examples of trades similar the Trade Diaries from
the October issue? There are many books on the market but it seems that a simple –Lawrence C.
compilation of trades containing hundreds of successful and losing trades would be
an excellent educational tool for people like me. If it doesn’t exist, you should do one! We have two online article indexes. The first,
Back to the markets I go. www.activetradermag.com/article_index.htm,
simply lists articles by year and magazine
–Stephen C. section. The Active Trader Store
(http://store.activetradermag.com) is more
Our trader wished to remain anonymous, but he informed us that he still needs tutelage and advanced — it has a search feature used to
experience himself. We don’t know of any publications that contain what the above writer find articles by name, subject, keyword,
described, but we will continue to publish Trade Diaries in all our magazines. author, and issue.
ACTIVE TRADER • January 2008 • www.activetradermag.com 5

OPENING Trades
International ETF roller
coaster: What’s in a range?
While the performance of many global
Tech hot, then not and regional ETFs in 2007 has been
Stocks take another blow quite impressive, their volatility has not
been for the feint of heart.
Stocks opened November — usually one of the More than 33 percent of the 49 “geo-
most bullish months of the year — with a sharp graphic” ETFs have gained more than 25
sell-off. The major U.S. indices shed anywhere percent, led by the Brazil index (EWZ),
from seven to 10 percent from the Oct. 31 high to which is up 77.72 percent (through Nov.
the Nov. 8-13 lows, prodded by continued fallout 9) and the FTSE/Xinhua China 25 index
from the sub-prime mortgage debacle, several big- (FXI; 63.3 percent).
time earnings misses, and stratospheric oil prices. However, those gains have not come
And the resurgent technology sector — still lag- Source: TradeStation without setbacks. More than 20 of these
ging the overall market since the 2000 tech-stock ETFs have had 2007 ranges (high price
crash — took things the hardest. The Nasdaq 100 (NDX) led the charge off the mid- minus low price, divided by low price)
August low, but it also fell a whopping 11.57 percent as of Nov. 12. However, when of more than 40 percent, with the
stocks rebounded intraday on Nov. 13, the NDX was again leading the pack. Golden Dragon Halter USX China index
(PGJ) joining EWZ and FXI with a triple-
Index Oct. 31 high Low Change digit range.
Russell 2000 830.39 762.38 (Nov. 8) -8.19% For comparison’s sake, the Nasdaq
S&P 500 1,552.76 1,438.53 (Nov. 12) -7.36% 100 index tracking stock (QQQQ) has
Nasdaq 100 2,239.23 1,980.18 (Nov. 12) -11.57% not had a range of more than 30 percent
Dow 13,962.53 12,975.11 (Nov. 13) -7.07% in the past four years, including 2007.
The FXI’s 2007 range of 144.58 percent
is on par with QQQQ’s 2001 range of
Third-quarter earnings for S&P 500 companies appeared strong at first glance, but a 154.14 percent. In 2006, FXI’s range was
closer look revealed some problems beneath the surface. 81 percent, although in 2005 it was only
Median profits grew 11.4 percent year-over-year on an earnings-per-share basis — 27 percent.
exceeding initial estimates of 8.5-percent growth and matching Q2 results, according In early March, EWZ, FXI, and PGJ all
to Charles Rotblut, senior market analyst at Zacks Investment Research. dropped at least 12 percent from the
However, many companies bought back shares during this period, a tactic that 2006 close.
inflated per-share profits. By contrast,
total net income for S&P 500 compa-
nies actually fell 26.9 percent in the E*Trade gets hammered
third quarter. And poor showings by
The sub-prime mortgage crisis hit online brokerage E*Trade hard on Nov. 12,
many marquee stocks — some
as the stock opened 36 percent lower and traded much of the day down 56
because of the sub-prime meltdown — percent from the previous close.
gave the quarter a negative pallor. E*Trade was one of the pioneers of the online brokerage revolution, and as
Analysts are now looking to 2008 the company (and its balance sheet) continued to grow, it branched off into
earnings with trepidation, because the other ventures such as banking, retirement planning, and mortgages.
full effects of last summer’s credit However, after management announced in early November that mortgage
crunch are still a mystery. “[No one] problems would result in much larger quarterly losses than originally expected
really knows how bad [the banks’] and refused to give earnings guidance for the rest of the year, E*Trade’s stock
losses will be,” says Rotblut. was downgraded by numerous analysts, with one saying the company had a
15-percent chance of going bankrupt.

Crude oil’s march to $100
Crude oil futures jumped 20.79 percent from the
end of September to Nov. 7, when the December
contract (CLZ07) peaked at a new all-time high
of $98.64. The advance represented crude’s
fastest six-week gain since April 2006. Source: TradeStation
Crude oil wasn’t alone at the top of the energy
heap, though. December gasoline futures Gold pulls back near
(RBZ07) rose 21.92 percent since the end of Source: eSignal historic level
September to trade at a two-year high of 2.4891.
Also, heating oil futures jumped 19.56 percent from the end of September to Nov. 8, Gold futures (GC) reached 847.50 — the
when they hit a record 2.6609. Natural gas gained 21.1 percent during this period to highest price since 1980 — on Nov. 8
reach an 11-month high of 8.712. before tumbling back toward 800.00 on
Nov. 12.
Treasuries keep rolling Who moved my cheese?

Ten-year T-note futures (TY) extended their up Influential dairy traders at the Chicago
move into November, topping 112-00 on Nov. 12 Mercantile Exchange are fighting the
— nearly nine full points above the mid-June low exchange’s decision to close the dairy pits
of 103-04. and make all trading electronic. The
For analysis of whether a full-fledged trend group sent a letter to the CME saying
change has occurred in the treasury market, see moving cheese and butter futures off the
“Bonds and the first rule of trading” on p. 22. Source: TradeStation floor could lead to price manipulation. It
also said the current, floor-based system
is working, and the CME needs to con-
Dressing for success sider what would happen if electronic
trading failed.
The details behind a sexual harassment dispute at hedge fund SAC Capital have raised The CME responded by praising its
some eyebrows. Former SAC trader Andrew Tong claims his boss Ping Jiang ordered regulation and surveillance systems,
him to take female hormones, with the alleged goal to improve Tong’s trading by mak- which it says are capable to monitoring
ing him less aggressive and more feminine. all markets.
After following Jiang’s directive, Tong claims he began wearing women’s clothes at
work and began a romantic relationship with his boss. Both SAC Capital and Jiang Pop Quiz:
Since 1986, how many non-overlap-
claim the accusations are false, according to CNBC and the New York Post.
ping eight-day declines of 10 per-
cent or more (high to low) has the
Nasdaq 100 index had? (Non-over-
Quoted: lapping means a day in one eight-
day period cannot belong to anoth-
“Analysts are good at research — they’re not good at er eight-day period.)
making recommendations or selecting stocks.” a. 63 c. 6

b. 24 d. 17
— Former Wall Street stock analyst Stephen McClellan (see p. 33). ending on Jan. 24, 2003.
Answer: 63, the last one

MARKET Pulse
The inside track

Riding the coattails of insider buyers might offer a trading edge for the rest of us.
BY DAVID BUKEY
S
everal days after
Caterpillar Inc. (CAT)
fell 8.9 percent
overnight on Oct. 20,
2005, Director Peter A. Magowan
bought $7.5 million of Caterpillar
stock. Magowan bought this large
stake when CAT traded just above its
three-month low. The stock jumped
51.65 percent within six months. FIGURE 1: GE INSIDER BUYING IN 2007 GE fell the first week after insiders reported stock
This purchase wasn’t confidential, purchases in 2007, but the stock bounced back over longer-term periods.
because the Securities and Exchange Source: eSignal
Commission (SEC) requires “insid-
ers” — all officers and directors of pub- February. But he also sold another $348 cy still exists, the following study gauges
licly traded companies — to report any million in July, just before it soared 31 how the 30 stocks in the Dow Jones
trades in their stock. Magowan reported percent within six months. Clearly, Gates Industrial Average (DJIA) behaved up to
this transaction within two days of his didn’t sell his stock because he lost faith six months after insiders reported buying
initial purchase. in Microsoft. significant amounts of their stock since
But are insiders always good investors? Insider-buying reports are easier to April 2005.
Admittedly, this is a cherry-picked exam- interpret than insider-selling reports,
ple; not all insiders buy at such fortuitous but neither offers clear indications of Are insiders really buying?
times. However, when insiders buy their what might be motivating company Each time company insiders trade, they
stock directly, they expect it to climb. insiders. Blindly following an insider’s must file a “Form 4” with the SEC, which
By contrast, insiders sell their stock for lead is no better than taking a “hot” lists several details about the trade: type
countless reasons — to diversify their stock tip from a friend. (buy or sell), date, share amount, stock
holdings, pay bills, exercise options — Despite this caveat, several academic price, and so on. Form 4 includes a great
not necessarily because they think the studies have shown stocks tend to rally deal of information, including whether
company’s stock will tank. Bill Gates, for within 12 months after insider buying. insiders bought shares directly and if that
example, sold at least $570 million of However, much of this research is purchase was part of their total compen-
Microsoft (MSFT) as it fell 8.72 percent in decades old. To determine if this tenden- sation package.

Traders who study these reports often
search for insiders — especially the top
brass (CEOs, CFOs, COOs) — who have
bought a significant number of shares
directly on the open market. Ideally,
executives will use money from their
own pockets.
On the other hand, automatic, sched-
uled purchases, such as stock rewards,
are more arbitrary. Any purchase is prob-
ably part of a compensation package if it
includes a stock price of $0, or if the
document’s footnotes discuss different FIGURE 2: INSIDER BUYING REPORT DAY Dow stocks dropped an average of 0.07
percent on days insiders reported stock purchases to the SEC.
types of stock (converted, restricted, or
deferred) or options.
Methodology Understanding the tables

This study examined insider-buying
reports provided by Edgar-Online.com Tables 1, 3, and 4 summarize price behavior for
different scenarios. They show the average,
and SecForm4.com. This data is also
median, maximum, and minimum price
available free at the SEC’s Web site
changes from:
(www.sec.gov), and Edgar Online lists a
limited amount of free data at Yahoo 1. The report day’s opening price to its
Finance (http://finance.yahoo.com). highest high on the same day (Table 1).
Significant insider buys were
2. The report day’s opening price to its
defined as:
lowest low on the same day (Table 1).
1. Purchases of $50,000 or more. 3. The prior closing price to its highest

2. Open-market purchases that high (largest up move, or “LUM,”) on
weren’t included in an executive’s any day between the 10th day after
total compensation, listed as “Buy” reported insider purchases and two
months later (Table 3) or the subse-
at Edgar-Online.com.
quent four months (Table 4).
3. “Direct” purchases that benefit only
that executive — not another firm, 4. The prior closing price to its lowest low (largest down move, or
trust, or foundation. “LDM,”) on any day between the 10th day after reported insider pur-
chases and two months later (Table 3) or the subsequent four months
The study focused on insider-buying (Table 4).
reports from the 30 Dow components
The standard deviations (StD) for the close-to-close changes are included, as
over the past two-and-a-half years — 98
well as the percentage of times the close-to-close change was positive (“%>0”).
transactions from 24 Dow stocks Figure A shows the close-to-close moves, LUMs, and LDMs from the initial bar to
between April 2005 to October 2007. the two subsequent bars.
The analysis excluded six Dow stocks
continued on p. 1 0

Market Pulse continued
98 instances Report day Open to low Open to high

Avg: -0.07% -0.82% 0.85%
Med: -0.06% -0.66% 0.60%
Max: 4.19% 0.00% 5.24% other Dow component. Although GE has
Min: -3.67% -4.68% 0.00% climbed 11.4 percent since April 2005,
StD: 1.19% 0.76% 0.84% the stock traded sideways much of the
Pct. > 0: 41.84% time, so few GE insiders earned quick
profits from these trades.
TABLE 1: REPORT-DAY STATS On the day insiders reported purchases, the stocks
dropped 58 percent of the time that day. The six significant insider buys in 2007
highlighted in Figure 1 (p. 8) give a
mixed picture. General Electric rallied
13.3 percent during this period, which
means all insiders have made money so
far, assuming they haven’t sold yet. But at
some point, each insider faced unrealized
losses in the first two weeks after buying.
Stocks drop on report day…

Figure 2 (p. 9) shows how stocks per-
formed on the day insiders reported pur-
chases to the SEC. The figure also com-
pares this daily close-to-close perform-
FIGURE 3: TWO WEEKS AFTER INSIDER BUYING Stocks lagged their benchmarks in ance to the average close-to-close move
the two weeks after insiders reported purchases. since January 2005 (“benchmark”). Stocks
fell an average of 0.07 percent on report
insiders either didn’t buy during this peri- Because traders can’t exploit any price days since April 2005 and lagged the
od, or which failed to meet the criteria: moves before insider buying is reported benchmark, which was a gain of 0.03
Merck (MRK), International Business (unless they already own shares), all pat- percent.
Machines (IBM), Exxon Mobil (XOM), terns shown here began on the day insid- Table 1 (p. 9) shows the statistics
Hewlett-Packard (HPQ), Proctor and ers submitted Form 4 to the SEC (report behind the report day’s loss. (For a
Gamble (PG), and Altria Group (MO). day). detailed explanation, see “Understanding
Insiders typically reported their pur- the tables.” For a discussion of the
chases to the SEC one or two days after Insider buying in General Electric difference between average and median,
buying shares. But the lag between trans- Insider buying at General Electric (GE) see “Key concepts,” p. 64.) These
action and filing dates varied from the met our requirements 27 times during the numbers also point to a loss: Although
same day to more than a month later. analysis period — far more than any individual stocks’ open-to-low moves are
98 instances Day 1 Day 2 Day 3 Day 4 Day 5 Day 6 Day 7 Day 8 Day 9 Day 10
Avg: -0.06% -0.06% 0.02% 0.14% 0.13% -0.18% 0.10% 0.14% 0.01% -0.01%
Med: -0.07% 0.03% 0.14% 0.14% 0.02% -0.09% -0.06% 0.03% 0.00% -0.11%
Max: 2.98% 1.89% 4.60% 3.87% 3.87% 2.52% 4.47% 2.79% 2.94% 2.24%
Min: -3.94% -4.27% -4.30% -4.27% -2.74% -7.11% -1.99% -3.76% -4.30% -1.64%
StD: 1.24% 1.10% 1.38% 1.25% 1.09% 1.14% 1.10% 1.18% 1.09% 0.80%
Pct. > 0: 46.94% 51.02% 56.70% 52.58% 50.52% 41.24% 45.36% 50.52% 49.48% 43.75%
TABLE 2: POST-REPORT STATS These statistics reveal that stocks traded sideways in the two weeks after reported insider buying.

roughly identical to their open-to-high
moves, stocks closed lower 58 percent
of the time.
…and trade sideways in the next

two weeks
Figure 3 (p. 10) shows Dow stocks’
average daily gain or loss on each of the
next 10 days. It also compares the
cumulative average move after insider
purchases to the benchmark move (blue
and red lines, respectively).
The stocks went nowhere, lagging the FIGURE 4: THE NEXT TWO MONTHS Dow stocks climbed an average of 1.76
percent over the next two months. But the median gain of 0.84 percent
benchmarks in the two weeks after lagged the benchmark move during the analysis period.
insiders bought shares. They climbed
roughly 0.15 percent on days 4 and 5,
slid 0.18 percent on day 6, and then 98 instances Next two months LUM LDM
rebounded at least 0.10 percent on each Avg: 1.76% 5.68% -5.04%
of next two days. But the cumulative Med: 0.84% 4.59% -4.50%
move never exceeded 0.25 percent in Max: 17.18% 18.46% 0.00%
either direction. Min: -10.42% 0.00% -20.29%
Table 2 (p. 10) shows the statistics StD: 5.86% 4.27% 3.79%
Pct. > 0: 53.68%
from Figure 3 and adds evidence to this
argument. No single day (or multi-day TABLE 3: TWO-MONTH PERFORMANCE Although Dow stocks rose after insider buying,
period) stands out as especially bullish the gains were fairly weak: Stocks moved up just 54 percent of the time, and
or bearish. For instance, stocks neither LUMs were only slightly larger than LDMs.
climb nor drop more than 59 percent of
the time on any day in Table 2.
Longer-term periods:
Stocks climb, but lag benchmarks
Figure 4 shows Dow stocks’ average and
median moves in the next two months
after insider-buying reports — from the
close on day 10 to the close on day 50.
These values are compared to the aver-
age 40-day move of all analyzed stocks
since January 2005.
Stocks rallied an average of 1.76 per-
cent during this period, beating the
1.40-percent benchmark move. FIGURE 5: LONGER-TERM MOVES Dow stocks jumped an average of 3.61 per-
cent from the second to the sixth month after insider buying. But again,
However, the median gain of 0.84 per-
this move wasn’t much higher than the benchmark, and the median
cent is less than half as large as the aver- 1.51-percent gain was much smaller.
continued on p. 1 2

Market Pulse continued
Related reading
“The bandwagon trade”
Active Trader, March 2002. 98 instances Next four months LUM LDM
Novice traders often hop on “hot Avg: 3.61% 9.54% -6.08%
stocks” they hear about in the Med: 1.51% 6.91% -5.38%
media or from friends. This strategy Max: 30.28% 31.89% 0.00%
shows why doing so almost always Min: -19.67% 0.37% -20.59%
results in losses. StD: 9.45% 7.61% 4.41%
Pct. > 0: 53.41%
“Market Pulse: Stock Market
Patterns & Tendencies, Vol. 1” TABLE 4: LONG-TERM PERFORMANCE STATS Dow stocks were mixed over the next four
This collection of 12 past Active months. LUMs were higher than LDMs — a bullish sign, but stocks rose only 54
Trader articles provides in-depth percent of the time and lagged the benchmark moves (not shown).
analysis of the stock market behav-
ior surrounding a variety of price,
indicator, and volume patterns, but it shows stocks’ average and median lagged the benchmarks, so mimicking
including gaps, “reversal days,” VIX moves in the subsequent four months (80 insider buying probably isn’t a good idea.
swings, put/call ratios, TRIN signals,
days) — from the close on day 50 to the This research was somewhat limited in
and others. Detailed probabilities
from multiple years of testing are close on day 130. It tells a familiar story its exclusive focus on the 30 Dow compo-
included in each article. These arti- as stocks advanced but still lagged the nents over the past two-and-a-half years.
cles sidestep hype and show the benchmark. Stocks jumped 3.61 percent, Moreover, only 24 stocks were analyzed
good and the bad. on average, in the second to sixth month and nearly one-third of the examples
after insider buying, but the median gain involved General Electric. Examining
You can purchase and download
of 1.51 percent is much smaller than both more stocks over longer periods may
past articles at www.activetrader-
mag.com/purchase_articles.htm. the average and benchmark (3 percent). yield different results, especially if you
Again, the discrepancy between aver- studied different market phases such as
age and median moves suggests individ- the 1996-1999 bull market or the 2000-
age value, a discrepancy that implies a ual stocks’ typical gains after insider pur- 2003 bear market.
few big gains have skewed the average chases are less dramatic than they seem. Also, the study’s criteria aren’t set in
higher, which makes the median value The relationships among average, median, stone. Possible adjustments include larger
more representative of the typical price and benchmark gains in Figures 4 and 5 purchases ($100,000 or more), multiple
behavior. Although stocks jumped during are nearly identical, which suggests this buys from different insiders within a short
this period, it’s likely they didn’t outper- pattern lasts up to six months. amount of time, or focusing solely on
form the broader market. Table 4 lists the statistics behind Figure high-level executives.
Table 3 (p. 11) shows the details 5’s four-month move. The details are bull- Finally, some traders believe buying
behind Figure 4’s two-month move. At ish: LUMs are bigger than LDMs, and opportunities arise when an insider buys
first glance, the statistics are bullish: The stocks climbed more often than they at significant highs or lows. If, for
LUMs are larger than the LDMs and dropped. However, stocks gained ground instance, CEOs buy their stock at a 52-
stocks posted gains more often than loss- just 53 percent of time, and the 3-percent week high, they presumably believe it will
es. But this up move wasn’t particularly benchmark (see Figure 5) is a fairly high continue to rally. Also, if CEOs buy their
compelling — stocks climbed just 54 per- hurtle. stock at a 52-week low, they think it will
cent of the time, and the 0.84-percent bounce back.
median gain trailed the benchmark gain. Further research Although this analysis shows that
In short, the DJIA rose 32.62 percent dur- These insider-buying patterns failed to insider-buying reports aren’t viable trade
ing the analysis period, so this behavior uncover a promising trade idea. After signals in Dow stocks, it doesn’t mean
could largely be the result of stocks drift- insiders reported purchases, Dow stocks they aren’t useful. It may require another
ing higher with the overall market. initially traded flat before rallying over technique or different perspective to glean
Figure 5 (p. 11) is similar to Figure 4, the next six months. But those advances some tradable information.!

Trading STRATEGIES
Designing and testing

a pattern-based trading system
We expand a study of multi-day highs and lows into a promising short-term reversal system.
BY HOWARD BANDY
A recent article (“Market facts:

Three-day pivots,” Active
Trader, August 2007) ana-
lyzed 10 years of S&P 500
data and showed Monday was most often
a three-day pivot high between April 16,
1997 and April 11, 2007, while
system.
We begin by making two observations.
First, it’s impossible to tell that any day is
a three-day pivot until the market estab-
lishes an intraday range three days after
the pivot day. Second, we cannot buy the
is statistically significant. But those statis-
tics are irrelevant if the trading system is
not profitable, so let’s design and test the
trading system first.
Try several different ideas

low or sell the high. The first idea tests the use of three-day
Wednesday was most often a three-day However, we can make some adjust- pivot entry conditions. If the entry rule is
pivot low during this period. ments to be able to trade on estimations a good one, trades should be immediately
This article takes the analysis a little of these. A day cannot be a three-day profitable.
further and develops a technique for pivot low unless it is a four-day low. That
turning those observations into a trading is, the low of a three-day pivot low day is Long trades:
lower than the lowest low of the three 1. Enter long trades on Wednesday
previous days. This means any day that when today’s low is lower than the
has a low that is the lowest of four days lowest low of the previous three
has a possibility of becoming a three-day days.
pivot low. 2. Enter using a limit order to buy at
We can either buy at a limit at the low- the lowest low of the previous three
est low of the previous three days, or we days.
can buy at the close. We are certain to 3. Exit long trades at the close after
be filled in both cases. We try other three days.
entries, such as buying on a limit some
percentage below the lowest low of the Short trades:
previous three days, but then we could 1. Enter short trades on Monday when
not be certain of a fill. today’s high is higher than the
We used daily data for the S&P 500 highest high of the previous three
Three-day pivot Depositary Receipts (SPY), the exchange- days.
traded fund that tracks the S&P 500 2. Enter using a limit order to sell
The Market Facts column in the index, from Jan. 1, 1995 to Sept. 21, short at the highest high of the
August 2007 issue of Active Trader 2007. The data is divided into an “in- previous three days.
defined a three-day high pivot as a sample” data set (Jan. 1, 1995 through 3. Exit short trades at the close after
day where the high is higher than Jan. 1, 2005), which is used to develop three days.
the highest high three days before the model; and an “out-of-sample” data
and three days after. Similarly, a set (Jan. 1, 2005 through Sept. 21, 2007) (The AmiBroker code for these rules is
three-day low pivot is a day where that is used to test it. listed in “System code” — ThreeDay.afl
the low is lower than the lowest Some statistical analysis could deter- on p. 15.)
low three days before and three mine how often four-day lows become The results are promising. From Jan. 1,
days after. three-day pivot lows, on what days of the 2005 to Sept. 21, 2007, long trades
week this happens most, and whether it gained 16 percent per year while exposed

Analysis definitions
Correlation: A measurement of the similarity of Maximum System Percentage Drawdown (MaxDD):
movement between two series. A value of -1.00 means The largest peak-to-valley percentage decline in portfolio
that the two always move opposite to each other, a equity. Lower values are preferred.
value of +1.00 means the two always move in the same
direction and by the same amount, and a value of Out-of-sample: Data that has not been used in any
0.00 means that the two move independently. way to develop the model. Out-of-sample data can be
from the same series (typically, price data), but later;
In-sample: The data that is used to develop the model. or it can be from a different series, provided the series
The results of running a trading system over in-sample is uncorrelated to the in-sample data. When a trading
data are always good because the system is tweaked system is tested over out-of-sample data, the results are
until the results are good. Performance over in-sample used to estimate the likelihood that the system will be
data has little or no value in predicting the likelihood a profitable when traded with real money.
trading system will be profitable when traded with real
money. Risk Adjusted Return (RAR): As used in this article,
and in the reports generated by AmiBroker, RAR is the
K-Ratio: The linear regression of the slope of the equity compounded annual rate of return as a percentage,
curve, divided by the standard error of the equity, divided by the percentage of time the system is holding
normalized. Higher values are preferred. (See "Key a position. Higher values are preferred.
concepts," p. 64, for greater detail.)
18 percent of the time — a net risk-

adjusted return (RAR) of 89 percent and 1/1/1995 to 1/1/2005 1/1/2005 to 9/21/2007
an annual RAR of 32 percent per year. RAR MaxDD K-ratio RAR MaxDD K-ratio
Short trades earned 0.5 percent while SPY 33% 12% 0.05 16% 4% 0.08
exposed 18 percent of the time for an OEX 9% 26% 0.02 16% 5% 0.09
annual RAR of 1 percent. For all trades, NDX 21% 33% 0.03 27% 6% 0.07
the system climbed 16 percent while COMP -2% 57% 0.00 19% 6% 0.06
exposed 35 percent of the time — an XLB -6% 34% 0.00 0% 19% 0.01
annual RAR of 16 percent. (See “Analysis XLE -8% 36% 0.00 2% 18% -0.01
definitions” for definitions of RAR, XLF 53% 21% 0.06 30% 5% 0.04
MaxDD, and K-ratio.) XLI 22% 18% 0.05 14% 7% 0.03
Although no optimization was per- XLK 51% 23% 0.07 5% 8% 0.01
formed before testing this system, the arti- XLP 3% 21% -0.01 12% 7% 0.05
cle that inspired it was certainly the result XLU 25% 17% 0.08 19% 13% 0.05
of data mining. Therefore, these trading XLV 35% 10% 0.07 28% 14% 0.05
results should be treated as in-sample XLY 21% 27% 0.04 8% 16% 0.00
results — and because in-sample results Russell 1000 Stocks -39% 70% -0.03 -7% 28% -0.01
are unreliable in regard to future prof- Russell 2000 Stocks -41% 52% -0.01 5% 32% 0.00
itability, validation is advisable.
TABLE 1: OUT-OF-SAMPLE TEST RESULTS The three-day system performed well on
Validation is usually done by testing sector ETFs and indices, but not on individual stocks.
the system on out-of-sample data — data Source: AmiBroker, Quotes Plus
that has not been used during its devel-
opment. Since there is no period of time
that has not been used in developing the period from Jan. 1, 1995 to Jan. 1, 2005 Look more closely at the results for the
system, data from other exchange-traded is reported separately from the period Nasdaq 100 (NDX) and the Nasdaq
funds (ETFs), indices, and stocks will be Jan. 1, 2005 to Sept. 21, 2007. The tests Composite (COMP). Although both tech-
used as the out-of-sample data. included both long and short positions, nology indices are closely related, the
Table 1 shows the results from testing but excluded commissions and slippage. results are quite different. Figure 1 shows
a variety of instruments from Jan. 1, 1995 The entries for Russell 1000 and a scatter chart with each data point repre-
through Sept. 21, 2007. Because of the Russell 2000 stocks represent an average senting the gain (or loss) for COMP on
high degree of correlation between most of tests run on all individual stocks that the horizontal axis and NDX on the verti-
stocks, that data is not truly out-of-sam- are components of those indices. As Table cal axis. Points that lie exactly on either
ple, even though they are different sym- 1 shows, the system works fairly well for axis are trades that were triggered on one
bols. To be consistent with results and to sector ETFs, but does not work well for issue but not on the other. The correla-
individual stocks. continued on p. 15
give an idea of recent performance, the

Trading Strategies continued
tion is very strong with an r-squared

value of 0.74.
However, the equity curves for the
two indices are very different, as shown
in Figure 2. Trading the Nasdaq
Composite index for those 12.75 years
resulted in a final equity of about 1.1
times initial capital — a gain of about 10
percent, with a drawdown of over 50
percent. By contrast, trading the Nasdaq
100 index for that same period resulted
in final equity about 2.9 times initial
capital, with a drawdown of about 30
percent.
Expanding the system

The system has several parameters: how
many days to look back when determin-
ing the multi-day high or low, whether FIGURE 1: NASDAQ 100 VS. NASDAQ COMPOSITE System results for the Nasdaq 100
to take trades at the close or at a limit, and Composite indices are strongly correlated (vertical and horizontal axes,
how long to hold positions, and whether respectively), but their overall performance (Figure 2) differs.
the day of the week matters. A more Source: AmiBroker, Quotes Plus
System code
// BuyMultiDayLow.afl
You can copy this code at // // SellMultiDayHigh.afl
www.activetradermag.com/code.htm // AmiBroker code for a trading system //
// that buys when the low of the day // AmiBroker code for a trading system
// ThreeDay.afl // is lower than a multi-day low. // that sells when the high of the day
// // // is higher than a multi-day high.
// Buy when the low is lower than the // BPSwitch chooses the BuyPrice to be either //
// lowest low of the previous three days. // Close or PriorLow // SPSwitch chooses the ShortPrice to be either
// Short when the high is higher than the // 0 == Prior Low; 1 == Close // Close or PriorHigh
// highest high of the previous three days. BPSwitch = Optimize(“BPSwitch”,0,0,1,1); // 0 == PriorHigh; 1 == Close
// SPSwitch = Optimize(“SPSwitch”,1,0,1,1);
// Long positions are entered using a // LLBBars determines how many days in the
// limit order at the previous lowest low. // long multi-day look-back // SLBBars determines how many days in the
// Short positions are entered using a LLBBars = Optimize(“LLBBars”,3,1,10,1); // short multi-day look-back
// limit order at the previous highest high. SLBBars = Optimize(“SLBBars”,10,1,10,1);
// // HoldDays determines how many days to
// Exit all positions at the Close after // hold the long position // SHoldDays determines how many days to
// three days. LHoldDays = Optimize(“LHoldDays”,1,0,10,1); // hold the short position
SHoldDays = Optimize(“SHoldDays”,3,0,10,1);
SetTradeDelays(0,0,0,0); // BuyDOW chooses which day of the week to
buy. // SellDOW chooses which day of the week to
PriorLow = Ref(LLV(L,3),-1); // DayOfWeek() == 1 on Monday, 2 on sell.
PriorHigh = Ref(HHV(H,3),-1); Tuesday, ... // DayOfWeek() == 1 on Monday, 2 on
BuyDOW = Optimize(“BuyDOW”,2,1,5,1); Tuesday, ...
BuyPrice = Min(PriorLow, Open); SellDOW = Optimize(“SellDOW”,2,1,5,1);
SellPrice = Close;
ShortPrice = Max(PriorHigh, Open); SetTradeDelays(0,0,0,0); SetTradeDelays(0,0,0,0);
CoverPrice = Close; SellPrice = C; CoverPrice = C;
FourDayLow = L<PriorLow; PriorLow = Ref(LLV(L,LLBBars),-1); PriorHigh = Ref(HHV(H,SLBBars),-1);

FourDayHigh = H>PriorHigh;
BuyPrice = IIf(BPSwitch==1,C,PriorLow); ShortPrice = IIf(SPSwitch==1,C,PriorHigh);
Buy = FourDayLow AND DayOfWeek()==3; //
Wednesday MultiDayLow = L<PriorLow; MultiDayHigh = H>PriorHigh;
Sell = BarsSince(Buy)>=3;
Buy = MultiDayLow AND Short = MultiDayHigh AND
Short = FourDayHigh AND DayOfWeek()==1; // DayOfWeek()==BuyDOW; DayOfWeek()==SellDOW;
Monday Sell = BarsSince(Buy)>=LHoldDays; Cover = BarsSince(Short)>=SHoldDays;
Cover = BarsSince(Short)>=3; **
**

general system will allow us to test those
questions and, at the same time, test the
robustness of the system.
Although you can combine the logic
for both long and short positions into a
single system, the process will be much
quicker if longs and shorts are handled
separately.
Going long
For long positions, we found the follow-
ing optimal results using the S&P 500
Depositary Receipts as the in-sample data
from Jan. 1, 1995 through Jan. 1, 2005
and using the K-ratio as the objective
function we wish to maximize (see
“System code” — BuyMultiDayLow.afl): FIGURE 2: COMPARISON OF EQUITY CURVES Despite a strong correlation, the Nasdaq
Composite’s equity curve gained just 10 percent since 1995 (blue line), while the
• Enter at: prior low Nasdaq 100’s equity climbed to roughly 290 percent during the same period.
• Length of look-back: four days Source: AmiBroker, Quotes Plus
• Holding period: four days
• Day of week: Wednesday
In-sample Out-of-sample
Table 2 shows the results of applying 1/1/1995 to 1/1/2005 1/1/2005 to 9/21/2007
these optimal values to daily SPY price RAR MaxDD K-ratio RAR MaxDD K-ratio
data in the in-sample and out-of-sample SPY 40% 14% 0.10 19% 5% 0.06
periods. It might be too restrictive to take
positions on only one day of the week. TABLE 2: THREE-DAY PIVOT SYSTEM RESULTS SPY’s performance dropped sharply in the
We removed that restriction, re-optimized out-of-sample period (2005-2007). These results indicate it might be too restric-
tive to enter long positions only on Wednesday.
the parameters, and found the following
Source: AmiBroker, Quotes Plus
results to be optimal:
• Enter at: close

• Length of look-back: nine days In-sample Out-of-sample
• Holding period: one day 1/1/1995 to 1/1/2005 1/1/2005 to 9/21/2007
• Day of week: any day RAR MaxDD K-ratio RAR MaxDD K-ratio
SPY 51% 12% 0.08 61% 4% 0.21
Table 3 shows how the SPY performed
TABLE 3: SYSTEM WITH RE-OPTIMIZED VALUES Both the in- and out-of-sample results
with the revised optimal values. Both the
are better when the system entered long trades on any day of the week.
in- and out-of-sample results are better
when there is no requirement to enter
trades on a particular day of the week.
Figure 3 shows the equity curve for columns (2005-2007) are truly out-of- we wish to maximize (see “System code”
SPY using the system that enters on any sample. Based on these statistics, buying — SellMultiDayHigh.afl):
day of the week. The strategy only enters multi-day lows looks promising and is
long positions from Jan. 1, 1995 through worth further development. • Enter at: close
Sept. 21, 2007. The darker line is the • Length of look-back: seven days
SPY, and the light blue line is this system’s Selling short • Holding period: four days
cumulative equity. The vertical line repre- Using the same process, we can test • Day of week: Thursday
sents the beginning of the out-of-sample whether a multi-day high is a viable sig-
period on Jan. 1, 2005. nal to enter short positions. We found Table 5 compares the test results from
Table 4 (p. 17) shows the test results of optimal results using SPY as in-sample the in-sample period to the out-of-sample
taking only long positions on various data from Jan. 1, 1995 through Jan. 1, period using these optimal values. Again,
symbols from Jan. 1, 1995 to Sept. 21, 2005, taking only short positions, and it might be too restrictive to take posi-
2007. The results in the final three using the K-ratio as the objective function continued on p. 17

FIGURE 3: EQUITY CURVE — LONGS ONLY, 1995-2007 Despite several minor pullbacks, the re-optimized system posted consistent
gains in SPY since 1995.
tions on only one day of the week.
1/1/1995 to 1/1/2005 1/1/2005 to 9/21/2007
Removing that restriction and re-opti-
RAR MaxDD K-ratio RAR MaxDD K-ratio
mizing, we found these results to be
SPY 51% 12% 0.08 61% 4% 0.21
optimal:
OEX 48% 4% 0.21
NDX 26% 6% 0.10
• Enter at: close
COMP 29% 4% 0.14
• Length of look-back: 10 days
XLB -8% 12% 0.01
• Holding period: one day
XLE 47% 9% 0.05
• Day of week: any day
XLF 47% 6% 0.08
XLI 20% 3% 0.06
We re-tested the system in both in-
XLK -19% 10% 0.00
and out-of-sample periods with those
XLP 31% 3% 0.10
optimal values (Table 6). The K-ratio val-
XLU 10% 5% 0.07
ues are considerably lower for the short-
XLV 13% 5% 0.06
only system than for the long-only sys-
XLY 49% 4% 0.21
tem. This suggests the long side is more
Russell 1000 Stocks 29% 28% -0.01
reliable and out-of-sample results for
Russell 2000 Stocks 41% 32% 0.00
short trades may be weak.
TABLE 4: OUT-OF-SAMPLE PERFORMANCE (LONGS) These out-of-sample statistics suggest Figure 4 shows SPY’s equity curve for
that buying multi-day lows has promise. short trades from Jan. 1, 1995 to Sept.
Source: AmiBroker, Quotes Plus 21, 2007. Table 7 shows the short sys-
tem’s test results for various ETFs,
In-sample Out-of-sample indices, and stocks since 1995. Again,
results in the final three columns (2005-
1/1/1995 to 1/1/2005 1/1/2005 to 9/21/2007
2007) are truly out-of-sample. But these
numbers suggest that selling multi-day
SPY 11% 9% 0.04 -1% 7% -0.01
highs is not profitable for this system.
TABLE 5: SHORT SYSTEM PERFORMANCE Restricting the system to selling short only
on Thursdays could explain its lackluster performance. Bottom line
Source: AmiBroker, Quotes Plus Multi-day lows and highs are useful indi-

FIGURE 4: EQUITY CURVE — SHORTS ONLY, 1995-2007 The short system tested on SPY (blue line) earned less than half as much as
the long system (Figure 3).
1/1/1995 to 1/1/2005 1/1/2005 to 9/21/2007
SPY 4% 15% 0.03 15% 3% 0.03 Related reading
TABLE 6: REVISED SYSTEM PERFORMANCE (SHORTS) Allowing the system to enter short Quantitative Trading Systems
trades on any day of the week improved performance. But these results show By Howard Bandy
that short trades are weaker than long ones. 2007, Blue Owl Press
Source: AmiBroker, Quotes Plus www.quantitativetradingsystems.com
Howard Bandy article:

cators of short-term market swings. to be profitable even in periods of declin- “Designing
However, buying multi-day lows is more ing prices.! a mean-reversion system”
robust than selling multi-day highs. In Active Trader, October 2007.
addition, buying multi-day lows appears For information on the author see p. 4. This in-depth analysis of the system
design process lays the groundwork
In-sample Out-of-sample for developing a robust swing-trading
approach.
1/1/1995 to 1/1/2005 1/1/2005 to 9/21/2007
RAR MaxDD K-ratio RAR MaxDD K-ratio Other articles:
SPY 4% 15% 0.03 15% 3% 0.03 “Testing the mean-reversion
OEX 2% 5% -0.01 z-score system”
NDX -4% 9% -0.02 Active Trader, November 2007.
COMP 13% 5% 0.05 Properly testing a trading system
XLB 0% 6% 0.03 brings it one step closer to
XLE -12% 16% 0.01 tradability.
XLF 19% 4% 0.09
XLI -2% 8% 0.00 “Market Facts:
XLK 1% 7% 0.01 three-day pivots”
Active Trader, August 2007.
XLP 1% 3% 0.00
Which days are most likely to be
XLU -2% 11% -0.01
short-term highs or lows in the
XLV 12% 6% 0.02 stock market?
XLY 24% 5% 0.07
Russell 1000 Stocks -5% 19% 0.00 You can purchase and download
Russell 2000 Stocks 5% 23% 0.00 past articles at www.activetrader-
mag.com/purchase_articles.htm.
TABLE 7: OUT-OF-SAMPLE PERFORMANCE (SHORTS) Both these out-of-sample statistics and
SPY’s equity curve (Figure 4) show that selling multi-day highs isn’t profitable.

Trading STRATEGIES
Symmetrical reversal bars

A bar that inverts the preceding bar is often thought of as a
reversal setup, but analysis shows if you take the pattern too literally
you might be getting set up.
BY ACTIVE TRADER STAFF
A n interesting type of price behavior occurs when a mar-

ket makes a price move during one trading period and
reverses the move in the next period.
An example is a market with an opening price in the upper
part of the day’s range and a closing price in the lower end of
Let’s quantify this pattern, find
more examples, and see what typical-
ly happens after them.
T-note pattern
FIGURE 1: SYMMETRICAL
REVERSAL The second
the day’s range, followed by a day whose high and low are very Reviewing 10-year T-note daily price bar essentially flips
near the previous day’s, with an open near the low and a close data over the past six years for com- the price action that
occurred in the first
near the high. One way to interpret this behavior is by inverting parable market formations led to the
bar.
the price action of the first day on the second day — it effective- continued on p. 20
ly negates the price momentum or trend
that was in effect on the first day.
For example, if the first day (which
opens near the high and closes near the
low) is interpreted by itself as bearish, the
second day (which opens near the low
and closes near the high) could be inter-
preted as bullish.
The bars labeled 1 and 2 in Figure 1
show an idealized example of this type of
pattern. The first bar opens in the upper
half of the bar’s range and closes in the
lower half; the second bar reverses that
pattern, opening low and closing high.
Price jumped higher the next day.
Figure 2 actually contains this pattern
(second highlighted pattern from the
right) and five similar formations in the
10-year T-note futures (TY) from 2005.
The first pattern was followed by a signifi-
cant up swing, the second was followed
by an even larger downtrend, the third FIGURE 2: MIXED SIGNALS This stretch of 10-year T-note futures prices from 2005
was followed by another down move, and contained several symmetrical reversal up patterns, which have lower opens
price action after the remaining three was and higher closes on the second bar. Price action after these patterns varied.
mixed. Source: TradeStation

FIGURE 3: SYMMETRICAL REVERSAL UP PATTERNS Price tended to

gain ground several days after these patterns, but only
after underperforming the first couple of days. Also, the per- FIGURE 4: SYMMETRICAL REVERSAL DOWN PATTERNS After these
centage of gains (%>0, red line) was better than 50 percent patterns, there was a high probability of lower prices for the
from day 3 to day 7 only. first three days, after which price tended to rise steadily.
following pattern definition, which will ABS = absolute value rical reversal pattern, the T-note initially
be referred to as symmetrical reversal Subscripts0,1.., denote today, slumped, closing essentially unchanged
bars: one day ago, etc. one day later and -0.4 lower after two
days — underperforming the slight gains
1. Today’s high and today’s low are Note: T-note prices trade in 32nds one would have captured by going long
both within 2/32nds of and halves of 32nds; prices have been on any given day. Also, the odds of a
yesterday’s high and low. converted to decimal equivalents for all higher close were less than 50 percent the
2. Yesterday’s open is in the top the calculations. (For T-note futures first two days after the pattern.
half of the day’s range and its pricing conventions, see “Treasury However, the pattern started to outper-
close is in the bottom half of refresher” in “Key concepts” on p. 64.) form the benchmark price movement on
the day’s range. Code can be copied from www.active- day 3, and by day 7 was significantly out-
3. Today’s open is in the bottom tradermag.com/code.htm. pacing typical gains — after which per-
half of the day’s range and its Twenty-one patterns matched this defi- formance turned down sharply. And, the
close is in the top half of the nition between May 2001 and October odds of a higher close was greater than
day’s range. 2007. The solid blue line in Figure 3 50 percent from day 3 to day 7.
shows the median close-to-close changes Overall, there might be some indica-
As simple formulas, the rules are: that typically occurred in the first 10 days tion of potential upside follow-through
after these patterns while the black line after these patterns, but this is hardly a
1. ABS(H1 - H0) < 0.0625 (“benchmark”) represents the correspon- strong case. At best there’s a delayed reac-
2. ABS(L1 - L0) < 0.0625 ding median price changes for all one- to tion after which (and only until day 7)
3. (O1-L1)/(H1-L1) > 0.50 10-day periods during the review period. the odds are a little better than 50 per-
4. (C1-L1)/(H1-L1) < 0.50 (See “Key concepts” for an explanation of cent the market will close higher than it
5. (O0-L0)/(H0-L0) < 0.50 the difference between median and aver- did the second day of the pattern.
6. (C0-L0)/(H0-L0) > 0.50 age.) This shows us how the post-pattern This pattern has an obvious counter-
price behavior compared to the market’s part: a day that opens in the bottom half
where random behavior. The red line shows the of the bar and closes in the top half and
percentage of times price closed above does the opposite the next day — the
L = low of the day the closing price of the pattern one to 10 implication being the market has reversed
H = high of the day days afterward. the previous day’s presumably bullish
C = close of the day Interestingly, after the two-bar symmet- sentiment:

FIGURE 5: RECENT PATTERNS Recent symmetrical reversal patterns
show the mixed behavior found in testing.
Source: TradeStation
the second was followed by mixed price

1. ABS(H1 - H0) < 0.0625 movement (manifesting the delayed reac-
2. ABS(L1 - L0) < 0.0625 tion implied by the analysis).
3. (O1-L1)/(H1-L1) < 0.50
4. (C1-L1)/(H1-L1) > 0.50 The next step
5. (O0-L0)/(H0-L0) > 0.50 Now that we’ve analyzed this data, it’s
6. (C0-L0)/(H0-L0) < 0.50 time to highlight a few points. First, the
analysis period is fairly long at six years,
There were only 17 patterns from May and it included the most current price
2001 to October 2007, bringing the total data at the time the analysis was conduct-
number of symmetrical reversal patterns ed. This means we really have no way to
(up and down) to 38. Did the market know how T-notes might behave in the
move lower after these symmetrical rever- future after these patterns. Also, we did
sal down days? not investigate the performance on small-
Figure 4 tells the tale. This time the er portions of data to see if the results
results are little stronger — at least for were different from one to the next.
the first three or four days after the pat- Also, the pattern definition itself could
tern. Price closes lower than the close of certainly be scrutinized: The opens and
the pattern for the first six days, and the closes could have been required to be
odds of a lower close are above 50 per- much nearer to the highs and lows; to
cent for these days (and 64 percent or increase the number of examples, the cri-
higher for the first four days). Notice, teria used here were as loose as possible.
though, the decline reached its maximum Finally, the price action leading up to
by day 3; after that, price is actually ris- each pattern was not included in the
ing. analysis.
Figure 5 shows two up patterns and These issues could dramatically change
one down pattern that occurred in the our understanding of these patterns, so
past six months. T-notes were in a strong they will be the subject of a follow-up
uptrend; the down pattern was a bust, article in next month’s issue of Active
while the first up pattern fared well and Trader.!

Advanced STRATEGIES
Bonds and the first rule of trading

All short-term trading strategies in bonds, notes and other markets affected by
long-term interest rates will need to be adapted to the reality of a new bond bear.
BY HOWARD L. SIMONS
T hrow a rock down LaSalle

Street, Wall Street or any
other thoroughfare infested
with trading types, and
chances are you will hit a trading system
designer. (And won’t that feel good?) The
chances are equally high that the systems
a daily view, and then onto various intra-
day segments.
The premise behind multiple-time-
frame analysis is simple: You have to
know whether you are in a bull or a bear
market. That is the first rule of trading,
markets different, they differ across com-
modities.
Bonds and notes

This brings us to the market at hand —
10-year T-notes. Ten-year T-notes were in
a bull market between 1981 and 2003,
and it goes beyond the dictum that in a
designer at some point in his life bull market you have to be long, bull- spectacular intermediate-term retrace-
preached the virtues of multiple-time- spread, or out. At the risk of inflaming ments in 1987, 1994, and 1999 notwith-
frame analysis: working your way down the more doctrinaire technicians, not only standing. The broad channel for yields,
from a monthly view to a weekly view to are the internal dynamics of bull and bear highlighted in magenta in Figure 1, was
lower. And critically, the market
continued to make lower highs all
the way into June 2007.
In 2003 and its aftermath, we
can draw a rising channel in
turquoise. With the prominent
exception of June 2007, the highs
were higher, and the lows were
higher too.
That lower high in June 2007
muddies the waters as to whether
we are in a bull or a bear market.
Let’s abandon the long-term chart
analysis and shift to a pair of fun-
damental indicators we can read
technically — “swaption” volatility
and the term structure of inflation
expectations.
Swaptions
FIGURE 1: THE LONG-TERM TREND IN 10-YEAR NOTE YIELDS Despite retracements in 1987, While futures traders think the
1994, and 1999, 10-year T-note prices were in a bull market between 1981 and 2003, world reflects active futures mar-
mirrored by the extended downtrend in yields shown here. From 2003 onward, the kets, this is a view as parochial as
picture is less clear. the famous Saul Steinberg New

A call swaption
gives the buyer
the right to receive
the swap’s fixed
rate of interest
and pay the
floating rate of
interest. This is FIGURE 2: SWAPTION VOLATILITY FELL DURING FLAT YIELD CURVE ERA Swaption volatilities
(plotted inversely) have been moving closely with swap rates since the yield curve
a bullish position began to flatten. Also, yields have tended to bottom at high swaption volatilities and
peak at low swaption volatilities — the notable exceptions occurring in the rate-hike
in bonds, as you era (vertical lines).
profit if rates fall

in the future.
Yorker cover summarizing the
New Yorker’s view of the world
west of the Hudson River. The
single-point maturity of a 10-year
T-note future is nice, but most
corporate financing terms are set
from the 10-year swap rate, which
is the present value of the yield
curve out to 10 years. (An inter-
est-rate swap is an agreement
between two parties to exchange
fixed and floating interest-rate
payments.) Those who are bullish
on the bond market elect to pay
the floating rate and receive the FIGURE 3: THE YIELD CURVE AND VOLATILITY ARE LINKED The more the two-year/10-year FRR
fixed rate around this swap level, exceeds 1.00, the steeper the yield curve. Higher volatility contributes to a steeper
and those who are bond bearish yield curve by widening the range of possible interest-rate outcomes, which forces
elect to pay the fixed rate and bond buyers to demand a higher yield.
receive the floating rate around
this swap level. rate of interest. This is a bullish position rise in the future.
A swaption is the right but not the in bonds as you profit if rates fall in the Swaption volatilities, plotted inversely,
obligation to enter into a swap at some future. A put swaption buyer has the have been moving closely with swap rates
point in the future. A call swaption gives right to receive the floating rate and since the yield curve began to flatten
the buyer the right to receive the swap’s pay the fixed rate; this is a bearish (Figure 2). And the pattern has been for
fixed rate of interest and pay the floating position in bonds as you profit if rates continued on p. 24

Advanced Strategies continued
yields to have bottomed at high swaption and swap rates stayed relatively low. This ship between the shape of the yield curve
volatilities and peaked at low swaption was the period when the yield curve was and swaption volatility (Figure 3). The
volatilities. flattening and inverting while the Federal yield curve is measured by the forward
A major divergence occurred between Reserve made 17 consecutive quarter- rate ratio (FRR) between two and 10
the two turquoise vertical lines: Swaption point rate increases. years; this is the rate at which you can
volatility plunged while both Treasury We can highlight the strong relation- lock in borrowing for eight years starting
two years from now, divided by the 10-
year rate itself. The more FRR exceeds
1.00, the steeper the yield curve.
Higher volatility contributes to a steep-
er yield curve as it widens the range of
possible interest rate outcomes and forces
bond buyers to demand a higher yield as
compensation. This relationship is so
strong that any increase in swaption
volatility will push the long end of the
yield curve higher and lead to higher
bond yields — lower prices for futures
traders used to thinking in these terms —
as a result.
Inflation expectations
The other determinant of the yield curve
capable of forcing the long end higher is
expected inflation. This was known for
years in theory but was difficult to prove
until the Treasury began issuing TIPS, or
FIGURE 4: EVOLUTION OF TIPS BREAKEVEN YIELD CURVE Over the past three years, the inflation-protected bonds in January
term structure of TIPS breakevens has oscillated between an inverted curve, 1997 (Figure 4).
implying declining future rates of inflation, and a positively sloped curve, imply- The difference between Treasury yields
ing rising future rates of inflation. For most of 2007, the curve has had a and TIPS yields is called the breakeven
strong positive slope. rate of inflation. This measure is surpris-
ingly impure. First, the Treasury rate can
plunge quickly in a financial panic as we
saw in February, August, and October
2007, amongst many other “flight-to-
quality” episodes. Second, TIPS yields are
buffeted by many embedded options,
including the tax rate on the accrual of
their principal and your own belief
whether the government does a good
(honest?) job in reporting inflation. But
these breakevens are the best we have to
work with, so work we shall.
Let’s construct a forward curve of TIPS
breakevens (Figure 5). Over the past
three years, the term structure of TIPS
breakevens has oscillated between an
inverted curve, implying declining future
rates of inflation, and a positively sloped
curve, implying rising future rates of
inflation. For most of 2007, this curve
FIGURE 5: THE FORWARD CURVE OF INFLATION EXPECTATIONS AND LONG-TERM YIELDS The FRR has been strongly positively sloping.
of expected inflation leads the 10-year Treasury rate by 96 days on average. Prior If we isolate one segment of this sur-
instances of rising FRR levels have led to increases in 10-year Treasury yields.
face that is between two and 10 years, we

can construct a FRR of inflation Related reading
breakevens parallel to the FRR for the Other Howard Simons articles:
Treasury yield curve itself. If we display
“Interest-rate shocks and currency moves”
the FRR as itself minus one and map it
Currency Trader, October 2007.
against 10-year Treasury yields, a striking
Short-term interest rates are typically cited as the prime catalyst of currency
pattern emerges: The FRR of expected moves. This study puts that idea to the test.
inflation leads the 10-year Treasury rate
“Stock shocks and the dollar”
by 96 days on average.
Currency Trader, September 2007.
Previous episodes of rising FRR levels,
Want to know what really happens to currencies after big stock market moves?
marked with green arrows in Figure 5,
have led to increases in 10-year Treasury “Currencies and Federal Reserve trade weights”
Currency Trader, July 2007.
yields. The configuration as observed in
The theory that a weaker dollar makes U.S. goods and services more
late October 2007 points to a similar competitive abroad sounds nice, but the facts argue otherwise.
episode of rising yields, and this is exactly
what we should expect. Rising inflation “Why currency traders should be humbler”
Currency Trader, May 2007.
expectations should lead to a steeper
A close look at the historical returns of professional currency traders is not for
yield curve, and unless it is offset by the feint of heart.
lower short-term rates, a steeper yield
“The stronger real: Don’t blame it on Rio”
curve is bearish for bonds.
Currency Trader, April 2007.
Lessons from past markets shed light on the possible future of Brazil’s
Resolving the conundrum high-flying currency.
The American economy has been import-
ing disinflation from China for the better “Comparing the major euro cross rates”
Currency Trader, March 2007.
part of this decade. Regardless of any of
Europe’s two major non-euro currencies — the British pound and the Swiss
the macroeconomic causes normally franc — reflect the growing new currency regime.
associated with rising inflation flashing
“Mexican peso: Who’s your padre?”
red, cheap imports made possible in part
Currency Trader, February 2007.
by a wildly undervalued Chinese yuan
The peso is one of several “emerging currencies” that have been gaining
have held back inflationary pressures in popularity. Find out about the key factor that has propped up the currency —
the U.S. which could disappear in a flash.
Moreover, the continued huge capital
“The new iron cross”
inflows from foreign investors mandated
Currency Trader, January 2007.
by the large and growing current account The long history of the D-mark/pound and now the euro/pound offers many
deficit have had the odd effect of keeping lessons about economic policies and currency fluctuations.
the yield curve flatter than it would have
“The pros make it look hard”
been otherwise. At some point, however,
Currency Trader, December 2006.
foreign bond buyers are going to demand Are currency traders making life unnecessarily difficult for themselves?
compensation against rising inflation
“Currency trends and volatility”
expectations, rising interest-rate volatility,
Currency Trader, November 2006.
and a weaker dollar. All three of those
Interesting insights come from putting currency volatility under a microscope.
factors will combine to push the long end
of the yield curve higher for any level of “Currencies and conventional U.S. investments”
Currency Trader, October 2006.
short-term interest rates.
The financial media often reports on moves in the stock and bond markets
The question of whether we stand at vis-à-vis currency fluctuations, but these relationships might not be what you
the end of a quarter century-long bull expect.
market in bonds or four years into a new
“The dollar index and ‘firm’ exchange rates”
bear market in bonds should be resolved
Currency Trader, December 2005.
in favor of the bear. All short-term trading The vast majority of currency traders are familiar only with the current
strategies in bonds and in markets affect- floating-rate system. But are we about to enter a new “firm exchange rate”
ed by long-term interest rates will need to era dominated by the dollar and euro?
be adjusted to accommodate this new
“Howard Simons: Advanced Currency Concepts, Vol. 1”
reality. Violate the first rule of trading and A discounted collection that includes many of the articles listed here.
rules 2 through N will not matter.!
You can purchase and download past articles at
www.activetradermag.com/purchase_articles.htm
For information on the author see p. 4.

TRADING System Lab
Reverse-trade
equity
management
BY VOLKER KNAPP
Market: Stocks.
System concept: Past Trading

System Lab articles have explored
“equity curve management,” the
object of which is to avoid taking
trades when a system’s equity curve
is trending down and resume trad-
ing when the curve turns back up.
The goal is to reduce risk and FIGURE 1: SAMPLE TRADE The equity curve management technique was applied to a basic
moving-average crossover strategy.
increase profitability by eliminating
Source: Wealth-Lab
likely losing trades during a system’s
drawdown phase.
“Trading the equity curve” (Active Trader,
December 2006) showed how this approach
turned an underperforming trend-following
system into a modestly profitable one. A
dual moving average crossover technique
was used to determine whether the system’s
equity curve was trending up or down.
When the short-term moving average of the
equity curve crossed below the long-term
average, trade signals were not taken,
although the system continued to be paper
traded (i.e., monitored without executing its
signals); when the short-term average
crossed above the long-term average, trading
resumed.
The approach used in this test is derived
from Joe Luisi’s Basic Money Management
chapter in the book Computerized Trading
(edited by Mark Jurik). It takes the technique
outlined in the previous paragraph one step
further: Instead of discontinuing trading FIGURE 2: EQUITY CURVE (BASIC SYSTEM) The original system was profitable
when the equity curve management rules but it underperformed buy-and-hold by a wide margin, making it a
imply the next trade has a high probability of good candidate for experimentation with the equity-curve management
technique.
failure, the system executes the trade signal in
Source: Wealth-Lab
reverse — that is, sell when a buy signal

occurs and buy when a sell signals occurs.
The premise is that the potential losing trades
that would occur during a downturn in the
equity curve will be turned into winners if they
are reversed. Instead of simply sitting on the
sidelines, this approach seeks to extract extra
profits during a system’s down phase.
In this test, changes in the equity curve’s
trend will be defined by crossings of an 80-day
exponential moving average (EMA) that uses
Welles Wilder’s exponential smoothing tech-
nique (see “Key concepts” on p. 64). When the
equity curve is above its 80-day average, trade
signals will be executed in normal fashion; when
the equity curve is below the average, trade sig-
FIGURE 3: EQUITY CURVE (MANAGED SYSTEM) An improved equity curve came at
nals will be reversed.
the expense of reduced profits.
The system illustrating this equity manage- Source: Reports-Lab
ment approach is a basic moving average
crossover technique with one additional filter
rule. 10-day SMA of closing prices crosses above the
80-day SMA of closing prices.
Strategy rules:
Entry: Figure 1 shows a sample trade. We shouldn’t expect much
1. If the portfolio equity is equal to or above its 80-day from this basic trend-following system. The equity curve of the
EMA, enter long tomorrow at the market when system without the equity management rules shows the system
today’s 10-period simple moving average (SMA) is has been making money since the equity market turned back up
above today’s 80-period SMA, and today’s 80-period in 2003 (Figure 2). Its performance essentially mimics the shape
SMA is higher than it was five days ago. of the buy-and-hold equity curve (not shown), but with a much
2. If the portfolio equity is lower than its 80-day lower return.
Wilder’s moving average, enter short tomorrow at the The basic system’s net profit was only 146 percent vs. 280
market using the same rules in step 1. percent for buy-and-hold, and its annualized gain of 9.4 percent
was well below the 14.3-percent buy-and-hold figure. The basic
Exit: system had a hefty drawdown of -34.7 percent that spanned a
3. Close long position tomorrow at the market if the seemingly endless 1,775 days from 2000-2003.
10-day SMA of closing prices crosses under the These factors make the system a good candidate for improve-
80-day SMA of closing prices. ment with the new equity management technique.
4. Close short position tomorrow at the market if the continued on p. 28
BASIC STRATEGY SUMMARY

Equity Equity
Profitability Original mgmt Trade statistics Original mgmt
Net profit: $146,229.11 $87,604.14 No. trades: 193 167
Net profit: 146.23% 87.60% Win/loss: 37.31% 42.51%
Profit factor: 2.44 1.95 Avg. profit/loss: 5.34% 4.13%
Payoff ratio: 4.11 2.76 Avg. holding time (days): 55.84 65.21
Recovery factor: 2.30 3.19 Avg. profit (winners): 24.20% 19.08%
Exposure: 51.92% 49.87% Avg. hold time (winners): 113.46 105.69
Drawdown Avg. loss (losers): -5.88% -6.92%
Max. DD: -34.68% -17.46% Avg. hold time (losers): 21.56 35.27
Longest flat period: 1,775 days 543 days Max consec. win/loss: 6/10 9/14
Improvements highlighted in blue

Trading System Lab continued
Money management: Risk 1 percent of account equity

per position.
Starting equity: $100,000. Deduct $8 commission and

0.1 percent slippage per trade.
Test data: The system was tested on the Active Trader

Standard Stock Portfolio, which contains the following 17
stocks: Apple Computer (AAPL), Boeing (BA), Citibank (C),
Caterpillar (CAT), Cisco (CSCO), Disney (DIS), General
Motors (GM), Hewlett Packard (HPQ), International Business
Machines (IBM), Intel (INTC), International Paper (IP), J.P.
Morgan Chase (JPM), Coke (KO), Microsoft (MSFT),
Starbucks (SBUX), AT&T (T), and Wal-Mart (WMT). The test
FIGURE 4: ANNUAL RETURNS OF THE REVERSE SIGNALS The reverse
used dividend-adjusted data from Yahoo.
signals had only three winning years and was out of the mar-
ket four years.
Test period: October 1997 to September 2007. Source: Reports-Lab
Test results: Figure 3 shows the system’s performance with However, the larger problem was the original system had
the equity management rules — which ultimately reduced prof- missed an opportunity to capitalize on an ongoing uptrend. The
itability. Let’s see how it affected other aspects of performance. dark green areas in Figure 3 represent the system’s cash position,
In the beginning of the test period, the system went into a which was at its peak during the system’s largest historical draw-
moderate drawdown and the reversal system rules kicked into down. This means the system didn’t create enough reverse-trade
action — and turned out to be wrong. opportunities and was mostly out of the market.
PERIODIC RETURNS
% profitable Max consec. Max consec.
Avg. return Sharpe ratio Best return Worst return periods profitable unprofitable
equity equity equity equity equity equity equity

Orig. mgmt Orig. mgmt Orig. mgmt Orig. mgmt Orig. mgmt Orig. mgmt Orig. mgmt
Monthly 0.81% 0.57% 0.25 0.19 13.21% 10.88% -8.78% -8.18% 57.50 50.83 6 6 5 8
Quarterly 2.54% 1.80% 0.34 0.27 23.70% 20.81% -17.32% -14.50% 65.00 55.00 8 7 2 3
Annually 9.27% 6.38% 0.70 0.59 31.57% 27.47% -13.53% -10.56% 72.73 63.64 5 6 2 2
Improvements highlighted in blue
LEGEND
Net profit — Profit at end of test period, less commission. Exposure — ning trades. Avg. loss (losers) — The average loss for losing trades. Avg.
The area of the equity curve exposed to long or short positions, as hold time (losers) — The average holding time for losing trades. Max
opposed to cash. Profit factor — Gross profit divided by gross loss. consec. win/loss — The maximum number of consecutive winning and
Payoff ratio — Average profit of winning trades divided by average loss losing trades.
of losing trades. Recovery factor — Net profit divided by maximum Avg. return — The average percentage for the period. Sharpe ratio —
drawdown. Max. DD (%) — Largest percentage decline in equity. Longest Average return divided by standard deviation of returns (annualized).
flat period — Longest period, in days, the system is between two equity Best return — Best return for the period. Worst return — Worst return
highs. No. trades — Number of trades generated by the system. Win/loss for the period. Percentage profitable periods — The percentage of peri-
(%) — The percentage of trades that were profitable. Avg. profit — The ods that were profitable. Max consec. profitable — The largest number
average profit for all trades. Avg. hold time — The average holding peri- of consecutive profitable periods. Max consec. unprofitable — The
od for all trades. Avg. profit (winners) — The average profit for winning largest number of consecutive unprofitable periods.
trades. Avg. hold time (winners) — The average holding time for win-

FIGURE 5: DRAWDOWN COMPARISON One positive: The modified system’s drawdown length was a third as long and half as
deep as the original system’s.
Source: Reports-Lab
Figure 4 shows the equity management rules had only three will help clarify this issue. In this test, we verified the effect of
winning years (an absent bar means the reverse-trade system had modifying a malfunctioning system by means of equity-curve
no trades that year). However, the modified strategy had better management.!
reward-risk characteristics than the basic system — Figure 5
shows the revised drawdown was shorter and only about half For information on the author see p. 4.
the size of the original drawdown. Trading System Lab strategies are tested on a portfolio basis (unless otherwise
Finally, the equity management version of the system found noted) using Wealth-Lab Inc.’s testing platform. If you have a system you’d
itself in the red (with an 8.9-percent loss) after 47 of its eventual like to see tested, please send the trading and money-management rules to
167 total trades. Its initial breakaway had an opposite effect on editorial@activetradermag.com.
overall performance. And although the number of total trades Disclaimer: The Trading System Lab is intended for educational purposes
was reduced from the original 193 (most likely, some trades only to provide a perspective on different market concepts. It is not
were simply filtered out by the new system’s altered behavior), meant to recommend or promote any trading system or approach.
Traders are advised to do their own research and testing to determine
there was no meaningful reduction in market exposure time (52 the validity of a trading idea. Past performance does not guarantee
percent vs. 50 percent). Figure 6 shows the stocks that generat- future results; historical testing may not reflect a system’s behavior in real-
ed the majority of profits remained relatively constant. time trading.
Among the positive aspects of applying this
risk management technique were a more than
threefold reduction in drawdown duration,
improved reward/risk ratios (e.g., net profit to
maximum drawdown), and a much better-shaped
equity curve.
However, the reduction in profitability is a
strong argument against this technique in its
present state.
Bottom line: Despite a handful of improved

statistics, the reverse-trade equity management
rule ultimately had a negative impact on this sys-
tem. However, it’s difficult to know from a single
test if the equity management technique simply FIGURE 6: PROFIT DISTRIBUTION The equity-curve management rules (right) had
doesn’t work or if it’s incompatible with this type little impact on which stocks generated profits.
of system. Testing it on other types of strategies Source: Reports-Lab

On the MARKET
Why is the sky blue … and why do traders trade?

People trade certain stocks for a variety of reasons. But are they the right reasons?
Academic studies suggest many traders have inflated perceptions of their abilities,
and often use the wrong basis to select stocks.
M ultiple studies have

shown it’s difficult to
trade successfully. Very
few traders actually make money, and
even some that do barely show enough
and some do indeed buck the odds and
become successful.
So what separates the profitable traders
from the ones who eventually close up
shop and go back to their other jobs?
Berkeley, titled “Learning to be
Overconfident,” Gervais finds that over-
confidence generally tends to decrease as
a trader becomes more experienced.
He says a common belief is that “non-
profit to cover expenses. If you ask a group of professors who rational” traders will underperform
Yet there’s no shortage of new entrants have been studying traders for several rational traders and eventually operate
into the industry, each one with hopes of years, one of the big reasons is overconfi- only on the fringe of the marketplace, if
becoming the next George Soros. The dence. at all. However, Gervais says that is not
allure of financial freedom and the ability “People in general are overconfident,” necessarily the case.
to go to work each morning in your paja- says Brad Barber, professor of finance at “Nonrational traders earn higher
mas is a big draw for potential traders, the University of California-Davis expected profits than rational traders by
Graduate School of Management. “And bearing a disproportionate amount of the
when it comes to investing, they do it too risk they themselves create,” Gervais says.
aggressively. “We have found the most overconfident
“Many traders will see some new infor- and nonrational traders are not the poor-
The allure of financial mation about a stock in the Wall Street est traders. For any given level of learning
Journal or somewhere online and think bias and trading experience, it is success-
they have information that gives them an ful traders, though not necessarily the
freedom and the ability advantage. However, the returns for trad- most successful traders, who are the most
ing are not worth it.” overconfident. Overconfidence does not
to go to work each make traders wealthy, but the process of
In confidence becoming wealthy can make traders over-
Simon Gervais, associate professor of confident.”
morning in your pajamas finance at Duke’s Fuqua School of
Business, says traders enter the profession Looking for stocks
is a big draw for without any real knowledge of their abili- in all the wrong places
ties. It is only after a period of successes Barber says traders are also too eager to
and failures that the trader can assess his trade “media darlings” — stocks that have
potential traders, and ability. However, most new traders don’t been the subject of significant coverage,
have a realistic view. either on television or in the newspapers.
some do indeed buck “In assessing his ability, the trader To illustrate this concept, Barber refers
takes too much credit for his successes,” to Entremed (ENMD), a biotech compa-
Gervais says. “This leads to overconfi- ny. In May 1998, with ENMD trading
the odds and become dence.” around $12, a lengthy story appeared in
In a report co-authored with finance the Sunday edition of the New York Times
successful. professor Terrance Odean from Cal- continued on p. 31

On the market continued
that indicated Entremed had possibly series of data to be representative of the pattern was different. On a weekly basis,
developed a new cancer-curing drug. underlying population or distribution,” stocks that were heavily bought earned
ENMD opened at 83 the next day and Barber writes in the report. “Observing strong returns in the next week, while
closed at 51.81. Within three days, it was strong recent returns for a security, an stocks that were heavily sold performed
trading below 30, although it would be a investor might conclude that this security poorly the next week. This pattern holds
few years before it traded consistently is the type (or has become the type) of true for three or four weeks before revers-
near 12 again. security that generates strong returns. ing for several weeks.
However, the interesting thing to Thus, past performance is extrapolated to Barber believes these studies support
Barber and his colleagues was that this the future.” his theories of traders being overconfi-
news wasn’t news at all. More than five dent and trading even without an infor-
months earlier, the scientific journal Noisy trading mational advantage. The fact that stocks
Nature had reported the same thing, as Many traders are what Barber considers that are oversold in one year outperform
had the Times. “noise” traders — they are trading on those that are overbought shows that
However, the timing and the place- something other than information. In traders initially “follow the herd” and
ment of the most recent article sent
traders scurrying to their brokerages, try-
ing to get in on the action. “People expect small samples and short series
“The traders treated the story as a
major breakthrough,” Barber says. “There
was a huge price spike even though that
of data to be representative of the underlying
information was already out.”
And sometimes traders act on informa- population or distribution. Observing strong recent
tion that isn’t even pertinent to the stock
they are trading. When MCI, a wireless returns for a security, an investor might conclude
communications company that has since
declared bankruptcy and been bought
out by Verizon, was trading, it used the
that this security is the type (or has become
symbol MCIC.
There is an “MCI” ticker symbol, but it the type) of security that generates strong returns.
represents Mass Mutual Corporate
Investors, a closed-end investment com- Thus, past performance is extrapolated to
pany. Nonetheless, studies done by
Barber and his colleagues showed that
whenever there was news on the wireless
the future.”
company, it affected the stock of the –Brad Barber,
investment company. University of California-Davis Graduate School of Management
Barber calls this the “limited attention
bias.” In a May 2006 report entitled
“Systematic Noise,” Barber writes, other words, the knowledge they have have a tendency to choose stocks their
“[M]any individual investors cope with about a particular stock does not give peers are trading.
the challenge of sifting through thou- them an informational advantage over Over the course of a year, the stock
sands of potential purchases by consider- other traders. loses its luster as traders are not there to
ing only stocks that otherwise catch their Nonetheless, noise traders can have a affect its price, and it tends to return to
attention. Using news, extreme price big impact on stock prices. One of its “proper” position.
moves, and abnormally high trading vol- Barber’s studies, done in conjunction with In the short term, a stock remains
ume to identify when stocks may catch Odean, showed that stocks heavily pur- “hot” or “cold” for about three or four
investors’ attention … individual chased by individuals over the course of a weeks before it is ignored by traders and
investors are more likely to buy, rather year (i.e., there was a greater volume the price becomes normalized.
than to sell, attention-grabbing stocks.” increase among individual traders from In his report on noise traders, Barber
Another trader mindset is the “disposi- the previous year than among institu- says, “[I]t is unlikely that the trades of
tion effect.” This, Barber says, causes tions) underperformed stocks heavily individual investors are coordinated by
traders to concentrate on stocks that have sold by individuals by 4.4 percent in the the same factors that contribute to insti-
previously had strong returns. following year. tutional herding, such as principal agent
“People expect small samples and short However, in shorter time periods, the concerns, rational information cascades,

or a rational response to correlated infor- they behave as if their personal involve- ing stocks at a higher price than they sold
mation. Shared psychological biases ment can influence the outcome of them for because that would produce a
appear to contribute to coordinated trad- chance events (the illusion of control). counterfactual that produced a greater
ing. The influence of one individual Online investors have access to vast profit.
investor on asset prices is negligible. quantities of information, generally man- According to the report, “Repurchasing
However, we find that buying and selling age their own portfolios, and trade at the a stock at a higher price increases regret.
decisions of individuals are highly corre- click of a mouse. These aspects of online Even if the repurchased stock increases in
lated and they cumulate over time. Thus trading foster greater overconfidence.” price after it is purchased, the investor
individual investors, sometimes referred Another commonality the professors will always know that he would have
to as noise traders, do have the potential found involved trading stocks that had done better had he not sold it in the first
to affect asset prices because their noise is been previously traded. In “Once Burned, place.”
systematic.” Twice Shy: Naïve Learning, All of these studies lean toward a com-
Counterfactuals, and the Repurchase of mon theme: Traders trade stocks for myr-
The information illusion Stocks Previously Sold,” Barber and iad reasons — but not necessarily
The amount of information available on Odean found that traders tend to buy because they have researched the stock
the Internet adds to a trader’s belief that specific stocks again if they had a prof- and have an informational advantage over
he or she is better informed than fellow itable experience with them in the past. other traders.
traders. Likewise, they tended to avoid stocks
“There is unlimited information online, they had previously lost money on. Trader beware
and that sounds good to traders,” Barber “This behavior does not appear to be Barber’s latest study involves traders in
says. “But the only important information driven by superior information about Taiwan, where trading is very popular.
would be something that other traders these stocks since the investors do not Barber chose Taiwan because pertinent
didn’t have. Traders go to a site where earn superior returns on the previous data is more easily obtained there than it
they can sort and filter stock according to winners they repurchase,” Odean says. is in the U.S.
their criteria and they think it gives them “Counterfactuals” occur when someone “We’re able to get specific data about
an advantage, but anybody can do it.” compares an actual outcome to his men- who started the trade (an individual or an
The Internet has also made it easier for tal projection of what might have been. institution); that’s not available in the
people to trade. Barber and Odean co- Selling a stock produces the obvious U.S.,” Barber says. “Another difference is
authored a report that asks in its title, counterfactual of not selling a stock. If a that Taiwan is retail-dominated. About 70
“Online Investors: Do the Slow Die trader then buys that stock back, the percent of volume comes from individual
First?” Traders who previously phoned counterfactual moves closer to reality. traders.”
their broker to place orders but who However, the reality still differs from the But Barber says the portrait between
switched to online trading traded more counterfactual, most notably in how it traders from the two countries is similar
and were more speculative. The study affects a trader’s profit. — active traders in Taiwan underperform
showed that these traders, who beat the “If the stock is repurchased at a higher the market, just as is the case in the U.S.
market by 2 percent while still using a price than it was sold for, the investor is After years of studying traders, Barber
phone, lagged the market by more than 3 less wealthy than he might have been had doesn’t think trading is in the best inter-
percent annually after switching to online he not sold it to begin with; if the stock is ests of anybody looking for some fast
trading. repurchased at a lower price, he is money.
“Reductions in market frictions (lower wealthier than he would have been had “For the average person, I think it’s a
trading costs, improved execution speed, he not sold it at the higher price before bad thing,” he says. “Far more people
greater ease of access) don’t explain these repurchasing it at the lower price,” Odean believe they will succeed than actually
findings,” Odean says. “Overconfidence says. do. There are far too many people enter-
can explain the trading increase and per- The professors’ conclusion is that ing the business.”
formance reduction.” traders repurchase stocks that have Barber says that if somebody was con-
Barber and Odean found that traders decreased in value since they were sold sidering opening a restaurant but discov-
who switched from the phone to the because reality dominates the “never ered that the average restaurant loses
Internet typically did so after a very prof- sold” counterfactual — no matter what money, they would question getting into
itable period, which perhaps led to happens to the stock in the future, they the restaurant business.
greater overconfidence. can always tell themselves they are better “But they hear the same thing about
According to the report, “People also off than if they had never sold the stock trading, and they think they are the
become more overconfident when given at the higher price and then repurchased exception,” Barber says. “They think they
more information on which to base a it at the lower price. are better than anybody else, but that’s
forecast (the illusion of knowledge) and Likewise, they refrain from repurchas- usually not the case.”!

ACTIVE TRADER Interview
Behind the Wall with An industry insider explains
Stephen McClellan
why you don’t matter to
Wall Street — and what you
can do about it.
BY MARK ETZKORN
S tephen McClellan wants

you to be your own
stock analyst. He thinks
you can do it — correc-
tion, he thinks you should do it, given
what he knows about Wall Street and the
information most people rely on to make
investment decisions.
Until retiring in 2003, McClellan
spent 32 years as a highly regarded Wall
Street stock analyst — 19 consecutive
years on the Institutional Investor All-
American Research team and seven years
on the Wall Street Journal poll — who
plied much of his trade at Street behe-
moths Merrill Lynch and Salomon
Brothers.
If you’ve always suspected there was
something crooked at the intersection of
Wall and Broad streets — and who could
blame you after the fudged numbers and
analyst complicity that emerged in the
wake of the tech implosion in 2000 —
McClellan’s new book, Full of Bull
(Financial Times press, 2007), is pretty
much your worst fears realized.
In it McClellan recounts his three-
decade career as a Street analyst and how
his profession morphed from a relatively
© 2007, Ethan Pines low-profile wing of the financial industry
continued on p. 34

Active Trader Interview continued
technology analyst. After a half-dozen whether they can attract more funds and
“Analysts are good years he jumped ship to Salomon keep their jobs. Hence, the book’s subti-
Brothers, where he spent eight years as a tle: Do What Wall Street Does, Not What it
at research — they’re vice president. His final, and longest, Says, to Make Money in the Market.
stretch on the Street was his 18 years at Because they are not chained to the
not good at making Merrill Lynch, where he was a first vice performance treadmill, McClellan argues,
president. individual investors are free to engage in
recommendations In 1984 — just as the personal com- true long-term investing — i.e., holding
puting age was hitting its stride — he positions for several years instead of sev-
or selecting stocks.” published his first book, The Coming eral months. He includes extensive guide-
Computer Industry lines for identifying the appropriate
Shakeout, an analysis of stocks for such a portfolio, with criteria
the computer industry ranging from low P/E multiplies to the
and its prospects. habits of a company’s CEO.
Full of Bull will certain- The book contains a bit of name drop-
ly confirm the suspicions ping and anecdotes some readers may
of many cynical investors find distracting, but they are intended to
about the trustworthiness help spice up what might otherwise be a
of analyst recommenda- dry treatise on analyzing balance sheets.
tions and the behind-the- The major message, though, comes
scenes wheeling and through loud and clear.
dealing on the Street. As McClellan, who says he still invests his
McClellan describes it, own money, has a different perspective on
the game is played by an the Street now that he’s no longer on it.
insular community of “I’ve been retired for a few years,
institutional investors which has given me the ability to see
and traders, investment things through the eyes of an individual
banks, and large corpora- investor,” he says. “I’m not caught up in
tions — the audiences to the trees — I can step back and see the
whom brokerage analysts forest.”
© 2007, Ethan Pines like McClellan primarily And one thing he sees is that the Street
gear their research. Retail does not operate for the benefit of the
into an immensely influential and prof- investors — although they don’t know it individual investor.
itable — but also conflicted and mislead- — hardly enter the picture.
ing — mechanism that is almost inher- Which is precisely where McClellan AT: The bottom line is that the bro-
ently incapable of providing straightfor- sees the friction, as well as the opportuni- kerage analysts have their institution-
ward, useful information for individual ty, for individual investors. Aided by al clients in mind, correct? The ulti-
investors. Because, as McClellan writes, round-the-clock financial news coverage mate goal of the analyst’s work is to
that’s not really its job. that turns analysts into celebrities, get the institutions to do business
McClellan, who recently turned 65, investors take the typical analyst buy, through their brokerage and gener-
grew up in the Chicago area and was ini- hold, or sell recommendations literally, ate commissions.
tiated into the markets while in college when in fact they usually mean some- SM: Yes, that’s how they get paid from,
through a summer job at a two-man thing quite different to the institutional say, a Fidelity, which allocates commis-
stock brokerage that operated at the investors the analysts are really address- sions every quarter on a per-analyst basis
Midwest Stock Exchange (now the ing. In essence, individual investors think given how much research, help, atten-
Chicago Stock Exchange). After college the brokerage-firm analysts on TV are tion, and assistance the brokerage analyst
and a tour in the Navy, McClellan earned talking to them, when in reality the ana- gives that institution.
an MBA from George Washington lysts are directing their messages to the But the brokerage firm analyst is also
University while working as an industry institutional clients who, ultimately, pay key to the trading desk, and it’s still fairly
analyst at the U.S. Department of their salaries. important to the investment banking
Commerce (DOC). Meanwhile, McClellan says, behind the department — no company that I’m
His first Street job was at a boutique scenes the money managers on the Street aware of would ever have a brokerage
firm called Spencer Trask, where his are trading the market furiously, chained firm as their investment banker if the
experience at the DOC ultimately helped to the annual (and quarterly, and weekly, firm wasn’t covering the stock.
him gain a foothold as a computer and etc.) performance reviews that dictate

AT: Are retail investors such an after- or inept, and their opinions were superfi- “that’s a positive” or “that’s a negative,”
thought that firms don’t even consid- cial or incorrect. but what they traditionally haven’t been
er the potential bump they might get It was a disservice to investors, and I paid for — and therefore are not adept at
from fanning the flames of public think the price is still being paid. — is deciding whether this is a great
interest? Individuals are still misdirected by Wall stock to own now, or for how long, or
SM: Individual investors are so low on Street. They take the Street literally — whether it should be sold, and so forth.
the totem pole that the research isn’t ori- they believe a “hold” rating means “hold There’s a systemic way Wall Street
ented or directed toward them. The ana- the stock,” when in reality it means “sell makes it almost impossible for analysts to
lysts certainly never have any direct con- the stock,” and they believe the price tar- express objective opinions. The bulk of
tact with the retail brokers and the indi- gets on stocks are real, while in reality [the analysts’] audience is demanding
vidual private clients. they’re pretty specious. They take the rec- positive opinions, so they’re very biased
The research essentially gets repack- ommendations from broker-dealer in that direction. They’re also demanding
aged and dumbed down for [the retail] “emphasis lists,” and those lists have big-company research, so analysts have
audience, but they’re really a secondary proven to be no better than all the other traditionally overlooked and neglected
— or even tertiary — concern in terms of positively rated stocks in the system — small companies. These inadequacies
the major broker-dealers. and in some years, the lists underperform inhibit good, objective conclusions, rec-
the overall market. ommendations, and opinions.
AT: But if that’s the case, why do I think investors still pay too much
they bother doing anything for retail attention to Wall Street’s directives, AT: I can’t help but get the impres-
investors? Why not just give the skin- whereas insiders know better than to take sion you’re going to great lengths to
ny to the big institutional clients Wall Street literally. Insiders know how avoid saying, “This is corrupt.”
behind the scenes and not even the game is played, but individual SM: (laughing) Well, that’s a strong word.
bother going on TV and talking up investors have never had anyone before
the companies? decode the confusing, misleading Wall AT: I know. What’s a less-strong
SM: Analysts like to have high profiles, Street practices. word?
get on CNBC, get quoted in the Wall SM: It’s convoluted and there are many
Street Journal, and be well known because AT: If analysts aren’t supposed to be conflicts of interest. A big institutional
it gives them more clout with the big stock pickers, and if their price tar- investor, such as a big mutual fund, has
institutions. gets are not meant to be taken literal- one interest and a corporate executive at
ly, what is their role supposed to be? a company the analyst is covering has
AT: Was television really the thing SM: An analyst is expert in doing another interest. Also, the traders on the
that changed the game in terms of research on industry sectors and compa- [firm’s] trading desk are different, the
analysts being perceived as expert nies, summarizing that analysis and individual investor is different, and the
stock pickers with the ability to move research, and putting it in reports to pro- analyst is even catering to the media to
stocks just by talking about them? vide a better understanding of what’s hap- some degree. There are too many audi-
Did it take off in the 90s? pening in that industry as well as what’s ences and there are too many inert, built-
SM: That’s basically when it started. In going on operationally and directionally in biases.
the 80s the analysts were pretty much at a certain company. Analysts are good at
doing their job — researching and focus- doing research; they’re not good at mak- AT: You mention in the book, actual-
ing on institutions — but they certainly ing recommendations or stock selection. ly, that analysts do more marketing
weren’t moving the market, and they than research.
weren’t TV personalities and household AT: So they’re gathering information SM: Absolutely, and that’s what every-
names to individual investors. and providing data in the form of body on the outside is so appalled by: the
That all changed in the 90s, probably research, but they’re not necessarily fact that most of an analyst’s time is spent
for the worse, because it misled millions interpreting the information in the on the road meeting with big institutional
of investors. It certainly gave analysts a form of selecting stocks or forecast- investors, marketing and selling, and on
bad name when the market crashed and ing price moves? the telephone or the squawk box dis-
it turned out their research was synthetic SM: They might interpret it and say cussing the research rather than doing the
research.
“Investors pay too much attention to Wall AT: Let’s talk a little bit about the rec-
ommendation game — when buy
Street’s directives. Insiders know better than to really means sell, for example.
SM: It’s hilarious. There’s a great story
take Wall Street literally.” continued on p. 36

Active Trader Interview continued
that happened fairly recently. An analyst fooled by them. treadmill. They’re not just scrutinized
had an “overweight” rating on a stock. I every quarter — although that’s when the
don’t know what the stock price was — AT: One of the other examples you biggest evaluation occurs — but also
let’s say $40. His price target on the stock write about is when a stock has four every day, every week, and every month.
was $30, indicating he expected the stock or five sell ratings, it’s about the time The portfolio managers are absolutely
to go down to that level. However, he the stock is going to bottom out. Do on a one-year treadmill. They don’t invest
had an overweight rating on it. professional traders time their trades in stocks — they temporarily hold them
Looking further at the report revealed in a contrarian way based on these for short periods. They’ve been forced
the overweight rating meant the stock ratings? into being traders themselves. Mutual
would outperform the other companies in SM: Traders — at least the ones on Wall funds are just hand-tied and inhibited
that industry sector — they’d all go Street — are good for five minutes or two from being able to [be true long-term
down, but the others would go down hours or two days. They’re not looking at investors].
more than this particular stock. So his two months.
rating tells all the individual investors AT: And that’s where you think
who take these things pretty literally, the opportunity is for individuals
“Well, that must be a good stock to — a multi-year perspective, right?
buy, I’ll make some money in it. Let SM: That’s exactly right. Individuals
me put that in my portfolio.” And have the unique advantage, unlike
he’s really saying, “It’s an ‘outperform’ most of Wall Street, to be able to hold
all right — it’s only going to go down a stock for, say, three to six years. If it
25 percent.” doesn’t do much or goes down one
The more obvious example, which quarter it’s not going to kill them —
happens all the time, occurs when an while that would hurt a mutual fund’s
analyst with a buy rating on a stock performance, which is getting looked
moves his rating to a “hold” or a at every quarter. Individuals should
“neutral” rating, which is a very nega- take advantage of this.
tive signal. He’s actually viewing the
prospects of the stock and the com- AT: I have to say, though, when
pany as a sell, but prefers to transmit reading your book I thought it
this signal in code to save face. might be a daunting task for an
individual investor to do the kind
AT: You mention in the book that © 2007, Ethan Pines of research you outline, and it
going down even one notch at would certainly take some time to
any time is a bad sign. But yes, when most of the Street master. There seems to be a lot of
SM: Exactly. But many individual already has a negative sell rating on a subtlety and nuance to making judg-
investors might literally decide to hold on stock, they pretty much understand the ments about these companies. Do
to that stock in this situation, thinking, stock’s not going much lower, and they you think individuals can really do
“Gee, the firm still has a ‘hold’ rating on it are very much watching and anticipating this?
— that means I should continue to hold the first analyst to upgrade it to a neutral SM: Most individuals can’t do anywhere
it.” That’s how they are misled. or a buy. If there’s anyone left to put out a near the degree [of research] that I write
But “hold” can also be a roaring, sell rating, it wouldn’t impact the stock about in the book. But I lay everything
screaming buy recommendation. Say an because it’s already completely discounted out for them so they at least have an idea
analyst has a rare sell rating on a stock and in its price. of the full scope of [what it takes] to be
he moves it up to a “hold.” That’s a posi- But a move up from a sell rating when your own securities analyst. If they can
tive move. He’s thinking the outlook is get- the bulk of the Street is already negative do a quarter or even 15 percent of what I
ting better and the stock is not going to go can have a big positive impact on the lay out for them, they’ll be way ahead of
down any more — things are improving. stock. where they are now.
That means it’s a great time to be buying And I think when they come across
the stock, because his next move — six AT: Let’s follow up on time frame. new investment prospects, if they apply
months down the road — will be to move Have mutual funds — which most some of these things I focus on — the
the hold up to a buy rating. people would probably consider the quality of the company, whether it has a
The rating systems are absurd. Street prototypical investment operations — humble management attitude, listening to
professionals know how to interpret the abandoned long-term investing? the conference calls, and especially strate-
ratings and rating changes, but outsiders SM: Oh, absolutely. Mutual funds are gies about preserving capital, picking
— individual investors — are completely caught on the quarterly performance low-P/E-multiple stocks, and investing in

NYSE vs. Nasdaq stocks — they will be invested in them. market timing. So, yes, if you don’t have
way ahead of the game. These guidelines And of course, their success in the 90s the time or inclination to do serious
will help them avoid losing as much and the amount of money partners can investing on your own, [ETFs] are the
money as they’ve lost in the past and help make by starting up a hedge fund has way to go — they are probably preferable
them as long-term investors. exploded to the degree that they’re so to mutual funds. They don’t encompass
huge they need to play by the same rules. dividends, though, and almost half of the
AT: What do you think should total market returns stem from dividends.
change on the Street in terms of lev- AT: How likely do you think these But my book is really oriented to peo-
eling the playing field for individual changes are? Will it take a catastro- ple who are doing their own stock invest-
investors? phe of some kind to have the ing. There are millions of them, and they
SM: First of all, research has to be made reforms you talk about come into could do it a lot better. Indexes would
independent of broker-dealer firms on being? play a role, perhaps, but it’s not really
Wall Street. It has to be at an arm’s length SM: I think it’s going to be difficult to what I’m focusing on.
so analysts can conduct unbiased, objec- enact many of these reforms if it’s busi- There have been a few tell-all books
tive research and not feel beholden to ness as usual. The bursting of the 90s about the wild 90s, but there’s never been
certain audiences. Internet bubble put certain reforms in an investment advice book from someone
I also think there needs to be some place, but not anywhere near enough, with my background. In the past, almost
reform regarding corporate executives. and if we muddle through this decade, every one of them has been written by
They are culpable in regard to the biased few additional reforms are likely. outsiders — someone in the media, a
research that comes out of Wall Street. It will take quite a major catalyst to get columnist, a professor, and once in while
They distort and warp research. They some of these reforms done — either a a money manager from a mutual fund,
have a lot of ways to penalize analysts bad stock market, or other problems sur- like Peter Lynch.
who are not positively recommending facing in regards to research analysts. But nobody from the stock brokerage
their stocks — withholding access, and side — no analyst — has done it. I’m
so forth. AT: And you’re of the opinion that essentially divulging everything I’ve
Much of the buy-side — the hedge investors didn’t really learn the les- learned over the past 32 years on Wall
funds, for example — also need to be sons of the Bubble? Street, which the outside world doesn’t
reined in and have the same regulations SM: No, they really didn’t. They still take really [know about].
that apply to the brokerage firms and the “buy” ratings literally, they still watch
other parts of Wall Street. They are such Cramer on CNBC, they trade right and AT: Could you have written this book
big commission generators that they can left, they still buy stocks with almost infi- while you were still an analyst on the
twist the arms of analysts, and brokerage nite P/E multiples — the Googles of this Street?
firms bend over backwards for them. world — because they’re glamorous. SM: I think it would be impossible to
I also think the way they spread They pay very little attention to risk, val- write this book while active on Wall
rumors about stocks they have short uation, quality, and so forth. I don’t think Street. You can’t say, “You know, a broker-
positions in is very destructive. Brokerage enough lessons were learned, absolutely. age firm emphasis list doesn’t mean
firms cannot do that, but hedge funds are much,” or “a brokerage analyst’s stock rat-
able to. The reforms have to be not just AT: One conclusion a reader might ing is pretty meaningless” and still be an
on the brokerage side but also on the come to after reading your book is, analyst on the Street.
institutional investor side. because of all the misleading infor-
mation from the Street and conflicts AT: Do you know if you’ve peeved
AT: The number of hedge funds has of interest regarding individual any former associates?
exploded in recent years, but they stocks, it might be better to invest in SM: I think I will probably cause a bit of
weren’t born yesterday. Why haven’t the overall market using exchange a stir on Wall Street, but I tried to do
they been subject to the same regu- traded funds (ETFs) or even stock everything in a very arm’s length manner
lation? index futures. — I didn’t use specific names or refer-
SM: Because they used to be pools of SM: I do mention exchange traded funds ences and made everything generic to the
very high net worth individuals, and not in the book, which are not a bad way to whole Street.
so much public money. There weren’t as own a sector of the stock market. My Time will tell, because the book is just
many and they weren’t as big, so they only caveat is that people are going to try barely hitting the bookstores this week
didn’t have as much clout with Wall to engage in sector timing — be in a cer- (ending Nov. 3). I’ve talked to a few of my
Street brokerage firms or analysts. They tain sector for six months or a year and former associates and they’ve said,
were more private, I guess you’d say. Now then move to another sector. It is basical- “McClellan, you are right on.” They’re a
they have pension fund investors, univer- ly impossible to be consistently correct little amazed I’m telling it, but they’re
sity endowment funds, and public money with that form of fairly active trading and certainly in agreement.!

The Face of TRADING Trading setup
Hardware: PC with AMD Athlon XP
2500+processor, 1.83 GHz RAM,
Lifelong trader 1.2 GB hard drive.
Software: TC2000, Profit.net.

Internet connection: Cable.
Brokerage: Direct access.
Name: Carl Wyman

Age: 63
Lives and works in: Seattle, WA ticing medicine, much of it working as an for three to five days.
emergency room physician. However, He will place a stop-loss a tick under
C
Wyman was always trading in his spare the previous day’s low, but he doesn’t give
time, in options at first and then later his trades much leeway, noting that his
arl Wyman first discovered expanding into commodities. rule is “up or out.” He may exit a trade at
trading in 1957 when his He likens trading to medicine in that the day’s end if the stock fails to rally
middle school gave its stu- both follow a set pattern. sharply higher.
dents a brief lesson on the “There is a certain protocol you follow Wyman estimates that for every 20
stock market. if someone comes in with chest pains,” he trades, “about 12 or 13 are washes —
Soon afterwards, through an account says. “It’s the same in the markets. You small losers or small gains; a couple will
opened in his father’s name, he bought need a plan for all market conditions and see decent gains; and one will see a great
one share of Duplan for $12.25. you have to be able to implement that gain.”
“The broker picked it,” Wyman procedure.” He uses trailing stops and takes partial
remembers. “I knew nothing about it.” In 2000, Wyman retired from medicine profits along the way. For example, if the
The stock went up to $100 and split 2- and has been trading full-time since. market has rallied for seven to 10 days,
for-1, but the company eventually went he often closes out half of his positions
bankrupt, leaving Wyman with nothing Trading methodology: Wyman puts and tightens stops on other trades.
to show for his initial investment. on about 40 trades per week, trading
However, “it was an interesting game,” he individual stocks and ETFs. He may hold Most important lesson learned:
says. successful trades for a few days or even a “You can really screw yourself up if you
He continued pursuing this interest month, but he exits losers at the end of are not careful,” he says. “Anybody who
throughout his teenage years, reading var- the day. He is a trend and momentum has been in this business for a while is
ious books. The classic How I Made $2 trader and generally looks to enter trades covered with bruises and scars. It is a
Million In the Stock Market by Nicolas on pullbacks. constant battle. If you want to succeed
Darvas was significant for him. He didn’t Wyman studies daily charts and you need to get back up and figure out
have any funds to invest, but on summer searches for stocks that are moving up (or what you did wrong.”
vacations with his family, he would down) with strong momentum. He uses
always have a book of stock charts with the TC2000 software program to scan for Best trading book: Battle For
him. stocks with the largest percentage gain in Investment Survival, by Gerald Loeb.
Wyman earned a bachelor’s degree in the past six months.
chemistry, but he knew he didn’t want to “I’m always on the lookout for stocks When not trading: Wyman fishes,
spend his life in a laboratory. He also got that are up,” he says. sails, and coaches his two children’s
an MBA, focusing on the financial mar- For example, he monitors stocks that sports programs.
kets. For his master’s thesis, he developed are making new 52-week highs and
a model to predict the four-year cycles of watches them for three- to five-day Best thing about trading: “I can
the stock market. However, after business pullbacks with narrow range and light do it from my home,” Wyman says.
school, Wyman decided he wanted to do volume. He places his orders at night, “I can take a vacation whenever I want.
something more “significant” with his life after the markets are closed, entering I can do this in 20 minutes or I can sit at
and attended medical school. one tick above the previous day’s high the screen all day long. When the fishing
Wyman ultimately spent 25 years prac- in stocks that have been consolidating is good, I’m gone.”!

Global MARKETPLACE
GLOBAL STOCK INDICES
1-month 3-month 6-month 52-week 52-week Previous
Rank Country Index Current gain/loss gain/loss gain/loss high low rank
1 India BSE 30 19,289.83 8.53% 29.18% 38.98% 20,238.16 12,344.44 2
2 Hong Kong Hang Seng 29,708.93 6.75% 35.61% 42.17% 31,958.31 18,587.72 1
3 Australia All ordinaries 6,728.10 1.67% 11.88% 6.30% 6,873.20 5,303.40 7
4 Canada S&P/TSX composite 14,118.18 -0.81% 4.11% 1.82% 14,646.82 12,249.98 10
5 Germany Xetra Dax 7,799.62 -2.53% 3.81% 3.64% 8,151.57 6,195.81 5
6 France CAC 40 5,683.22 -2.74% 1.12% -6.39% 6,168.15 5,217.70 4
7 UK FTSE 100 6,385.10 -3.19% 1.21% -3.31% 6,754.10 5,821.70 9
8 Singapore Straits Times 3,673.01 -3.91% 11.24% 5.62% 3,906.16 2,734.96 3
9 US S&P 500 1,475.62 -5.26% -0.07% -2.24% 1,576.09 1,363.98 6
10 Japan Nikkei 225 16,096.68 -5.67% -4.88% -8.90% 18,300.39 15,262.10 8
Currencies and indices are ranked by their one-month gain/loss.
FOREX (VS. U.S. DOLLAR)

Current price 1 month 3-month 6-month 52-week 52-week Previous
Rank Currency vs. U.S. dollar gain/loss gain/loss gain/loss high low rank
1 Canadian dollar 1.0763 5.49% 13.56% 19.17% 1.0847 0.842 3
2 Brazilian real 0.5763 4.12% 9.42% 16.73% 0.5785 0.4372 2
3 Australian dollar 0.9239 2.88% 7.94% 12.41% 0.9347 0.7615 1
4 Euro 1.4514 2.64% 5.08% 6.74% 1.457 1.269 5
5 Swiss franc 0.8705 2.54% 3.23% 5.44% 0.8754 0.7948 6
6 Japanese yen 0.008729 2.07% 2.82% 4.87% 0.00896 0.00805 9
7 British pound 2.0845 2.06% 2.32% 4.60% 2.0906 1.8834 7
8 Singapore dollar 0.6903 1.59% 4.69% 4.72% 0.6916 0.6375 4
9 Hong Kong dollar 0.1288 -0.16% 0.86% 0.70% 0.129 0.1264 8
NON-U.S.-DOLLAR FOREX CROSS RATES

1-month 3-month 6-month 52-week 52-week Previous
Rank Currency pair Symbol Current gain/loss gain/loss gain/loss high low rank
1 Aussie $ / Pound AUD/GBP 0.4433 0.77% 5.50% 7.41% 0.4483 0.3932 2
2 Aussie $ / Yen AUD/JPY 105.854 0.76% 4.98% 7.14% 107.831 85.975 1
3 Euro / Yen EUR/JPY 166.284 0.53% 2.20% 1.77% 168.96 149.227 4
4 Aussie $ / Euro AUD/EUR 0.6366 0.17% 2.71% 5.24% 0.646 0.5727 3
5 Pound / Yen GBP/JPY 238.829 -0.06% -0.49% -0.29% 251.095 219.288 5
6 Pound / Euro GBP/EUR 1.4363 -0.65% -2.64% -2.11% 1.5296 1.4226 6
All data as of Nov. 7

ETF Snapshot
Date: Nov. 13
The following table summarizes the trading activity in the most actively traded exchange-traded funds. The information does NOT constitute trade
signals. It is intended only to provide a brief synopsis of each market’s liquidity, direction, and levels of momentum and volatility. See the legend for
explanations of the different fields.
1-year 10-day 20-day 60-day Volatility
Market Sym Sector Category Vol move move % Rank move % Rank move % Rank ratio/rank
Positive one-year performance
FTSE/Xinhua China 25 Index FXI Index iShares 6.62 M 98.49% -13.86% 86% -8.51% 57% 42.94% 83% .49 / 85%
Brazil EWZ Regional iShares 15.00 M 88.82% -0.70% 0% 7.58% 25% 53.48% 98% .30 / 25%
Malaysia EWM Regional iShares 3.01 M 51.93% 0.86% 14% 7.02% 60% 20.22% 89% .29 / 32%
Hong Kong EWH Regional iShares 8.42 M 48.91% -4.58% 67% 6.33% 32% 34.73% 98% .37 / 37%
Emerging Markets EEM Emerging Markets iShares 17.34 M 43.84% -4.93% 78% 0.45% 2% 26.82% 96% .46 / 55%
South Korea EWY Regional iShares 2.82 M 41.02% -3.54% 36% 1.59% 13% 18.84% 67% .59 / 92%
Australia EWA Regional iShares 2.28 M 39.48% -5.01% 71% -0.28% 0% 24.72% 95% .41 / 42%
Singapore EWS Regional iShares 4.23 M 39.26% -4.37% 50% -3.87% 44% 14.19% 77% .52 / 38%
United States Oil Fund USO Energy Fund 3.26 M 36.04% 2.00% 5% 5.90% 12% 32.48% 92% .19 / 18%
Oil Services OIH Energy HOLDRS 8.80 M 34.12% -1.00% 15% -8.14% 78% 9.48% 43% .54 / 88%
Gold GLD Metals streetTRACKS 8.29 M 28.38% 2.29% 5% 5.32% 36% 21.50% 93% .32 / 38%
Energy XLE Energy SPDR 23.32 M 27.90% -1.79% 62% -5.96% 69% 8.67% 43% .53 / 93%
Gold GDX Metals Market Vectors 2.19 M 23.91% -2.54% 0% 0.61% 0% 32.91% 94% .39 / 45%
Mexico EWW Regional iShares 4.02 M 22.51% -7.89% 88% -9.48% 79% 4.75% 22% .82 / 98%
Ultra Nasdaq 100** QLD Leveraged index ProShares 3.26 M 21.43% -13.20% 67% -9.79% 67% 16.14% 68% 1.00 / 100%
Materials XLB Materials SPDR 9.15 M 20.31% -4.21% 92% -3.33% 70% 8.89% 75% .59 / 75%
Taiwan EWT Regional iShares 8.17 M 19.80% -7.63% 83% -3.69% 55% 8.15% 50% .87 / 98%
Utilities XLU Utilities SPDR 5.33 M 15.74% -0.98% 0% 1.61% 23% 6.98% 80% .38 / 63%
Nasdaq 100 QQQQ Index PowerShares 177.91 M 15.50% -6.49% 67% -4.03% 64% 9.05% 72% .86 / 100%
EAFE* EFA Index iShares 6.92 M 15.37% -2.77% 78% -0.49% 6% 10.27% 90% .53 / 65%
Industrial XLI Industrial SPDR 5.08 M 14.60% -1.79% 38% -3.28% 59% 3.89% 39% .67 / 90%
Technology XLK Technology SPDR 4.45 M 12.62% -5.88% 71% -3.96% 55% 6.02% 47% 1.11 / 100%
Russell 2000 Growth Index IWO Index iShares 5.38 M 12.44% -3.59% 80% -3.86% 68% 2.80% 28% .68 / 92%
Russell 1000 Growth Index IWF Index iShares 3.02 M 12.14% -3.38% 82% -2.64% 63% 6.18% 57% .73 / 93%
Consumer Staples XLP Consumer SPDR 3.39 M 10.88% 0.32% 21% 1.07% 20% 6.27% 92% .35 / 38%
Dow Jones Industrial Average DIA Index Trust 15.64 M 8.85% -3.40% 72% -4.33% 79% 1.66% 20% .86 / 92%
S&P Midcap 400 Index MDY Index Trust 5.54 M 7.38% -2.84% 69% -3.37% 60% 2.95% 15% .66 / 98%
S&P 500 Index SPY Index Trust 201.60 M 6.06% -3.25% 81% -3.71% 73% 2.38% 32% .84 / 98%
S&P 500 Index IVV Index iShares 3.01 M 6.01% -3.26% 81% -3.83% 74% 2.35% 33% .84 / 98%
Russell 2000 Index IWM Index iShares 90.78 M 0.50% -3.38% 56% -4.15% 60% 0.04% 0% .62 / 93%
Negative one-year performance
S&P Home Building Index XHB Index streetTRACKS 4.42 M -39.43% -7.95% 60% -1.95% 12% -16.12% 34% .18 / 48%
UltraShort Nasdaq 100*** QID Leveraged inverse index ProShares 33.59 M -24.25% 13.43% 67% 8.23% 56% -17.38% 84% .72 / 100%
UltraShort Dow 30*** DXD Leveraged inverse index ProShares 3.82 M -14.66% 7.24% 78% 9.32% 79% -4.01% 23% .80 / 92%
Retail XRT Retail SPDR 3.03 M -12.27% -4.25% 40% -5.75% 70% -5.36% 26% .45 / 80%
Financial XLF Financial SPDR 74.48 M -11.95% -4.84% 35% -7.32% 74% -6.96% 51% .70 / 93%
UltraShort S&P 500*** SDS Leveraged inverse index ProShares 16.96 M -10.38% 6.33% 71% 8.02% 75% -5.37% 37% .78 / 97%
Dow Jones U.S. Real Estate IYR Real Estate iShares 5.65 M -9.79% -3.79% 26% -4.45% 63% -0.31% 2% .66 / 95%
Consumer Discretionary XLY Consumer SPDR 4.20 M -6.21% -4.49% 85% -5.93% 67% -1.83% 5% .53 / 92%
Semiconductor SMH Technology HOLDRS 10.46 M -5.87% -2.98% 20% -8.19% 59% -10.27% 92% .36 / 27%
Russell 2000 Value Index IWN Index iShares 2.69 M -4.81% -2.94% 40% -4.47% 68% -2.90% 16% .50 / 93%
Retail RTH Retail HOLDRS 5.50 M -0.62% -2.95% 47% -3.49% 53% -0.87% 5% .57 / 75%
Japan EWJ Regional iShares 17.00 M -0.37% -4.35% 80% -3.19% 50% 1.41% 74% .82 / 95%
UltraShort Russell 2000*** TWM Leveraged inverse index ProShares 4.85 M NA 6.24% 50% 8.19% 53% -2.96% 62% .61 / 87%
* Europe, Australasia, and the Far East ** Tracks twice the move of this index. *** Tracks twice the inverse, or opposite, of this index.
Legend from the close 60 days ago to today’s close. reading of 100 percent means the current read-
Vol: 30-day average daily volume, in thou- The “% Rank” fields for each time window (10- ing is larger than all the past readings, while a
sands (unless otherwise indicated). day moves, 20-day moves, etc.) show the per- reading of 0 percent means the current reading
centile rank of the most recent move to a certain is smaller than the previous readings. These fig-
OI: Open interest, in thousands (unless other-
number of the previous moves of the same size ures provide perspective for determining how
wise indicated).
and in the same direction. For example, the “% relatively large or small the most recent price
1-year move: The percentage price move from move is compared to past price moves.
Rank” for 10-day move shows how the most
the close one year ago (250 trading days) to
recent 10-day move compares to the past twen- Volatility ratio/rank: The ratio is the short-
today’s close.
ty 10-day moves; for the 20-day move, the “% term volatility (10-day standard deviation of
10-day move: The percentage price move Rank” field shows how the most recent 20-day prices) divided by the long-term volatility (100-
from the close 10 days ago to today’s close. move compares to the past sixty 20-day moves; day standard deviation of prices). The rank is
20-day move: The percentage price move for the 60-day move, the “% Rank” field shows the percentile rank of the volatility ratio over
from the close 20 days ago to today’s close. how the most recent 60-day move compares to the past 60 days.
60-day move: The percentage price move the past one-hundred-twenty 60-day moves. A

FUTURES & STOCKS Snapshot
Futures snapshot as of Nov. 12
The following tables summarize the trading activity in the most actively traded stocks and futures contracts. The information does NOT consti-
tute trade signals. It is intended only to provide a brief synopsis of each market’s liquidity, direction, and levels of momentum and volatility.
Volume figures are for the most-active contract month in a particular market and may not reflect total volume for all contract months.
For a more extensive futures snapshot, see Futures & Options Trader magazine (www.futuresandoptionstrader.com).
Note: Average volume and open-interest data includes both pit and side-by-side electronic contracts (where applicable).
Price activity for CME futures is based on pit-traded contracts, while price activity for CBOT futures is based on the highest-volume contract (pit or electronic).
10-day 20-day 60-day Volatility
Market E-sym Pit sym Exch Vol OI move % Rank move % Rank move % Rank ratio/rank
S&P 500 E-Mini ES CME 1.87 M 1.97 M -6.90% 100% -7.69% 100% -0.67% 8% .85 / 98%
10-yr. T-note ZN TY CBOT 1.21 M 2.39 M 1.00% 53% 2.88% 94% 2.81% 27% .30 / 60%
5-yr. T-note ZF FV CBOT 599.8 1.68 M 1.62% 89% 2.71% 98% 2.58% 47% .35 / 73%
Nasdaq 100 E-Mini NQ CME 429.0 422.5 -10.20% 100% -8.77% 100% 5.14% 25% .84 / 100%
Eurodollar* GE ED CME 355.3 1.30 M -0.05% 0% 0.23% 38% 0.40% 45% .18 / 3%
30-yr. T-bond ZB US CBOT 344.1 951.3 0.89% 31% 3.70% 91% 3.66% 48% .26 / 27%
Crude oil CL NYMEX 289.2 313.5 1.17% 0% 9.86% 53% 31.75% 97% .23 / 28%
Russell 2000 E-Mini ER CME 246.5 585.9 -7.36% 100% -8.61% 96% -3.10% 30% .69 / 100%
2-yr. T-note ZT TU CBOT 233.4 1.01 M 0.83% 100% 1.06% 96% 1.04% 27% .34 / 65%
Mini Dow YM CBOT 187.2 101.8 -6.59% 100% -7.65% 97% -0.88% 26% .90 / 95%
Eurocurrency 6E EC CME 171.2 207.4 0.88% 22% 2.43% 48% 8.52% 99% .26 / 28%
Gold 100 oz. GC NYMEX 109.3 307.8 1.91% 5% 5.97% 40% 22.21% 96% .33 / 40%
Corn ZC C CBOT 102.8 542.6 0.80% 5% 4.70% 32% 15.41% 77% .33 / 47%
Soybeans ZS S CBOT 69.9 139.4 3.48% 56% 6.00% 56% 28.94% 99% .29 / 35%
Natural gas NG NYMEX 63.6 84.3 -0.16% 0% 6.93% 15% 31.80% 98% .43 / 55%
Wheat ZW W CBOT 47.9 194.3 -8.14% 63% -8.69% 50% 13.24% 10% .16 / 20%
S&P MidCap 400 E-Mini ME CME 30.1 102.5 -6.17% 100% -7.38% 90% 0.76% 3% .63 / 100%
Fed Funds ZQ FF CBOT 22.9 105.4 0.08% 38% 0.34% 96% 0.64% 92% .12 / 25%
*Average volume and open interest based on highest-volume contract (December 2008).
Stocks snapshot as of Nov. 13
1-year 10-day % 20-day % 60-day % Volatility
Stock Sym Vol move move rank move rank move rank ratio/rank
Companhia Vale do Rio Doce RIO 26.03 M 166.21% -3.32% 63% 3.01% 6% 74.90% 98% .25 / 20%
Research in Motion RIMM 28.86 M 161.72% -7.14% 67% 2.23% 0% 43.22% 37% .44 / 82%
Apple Inc. AAPL 39.43 M 99.95% -9.11% 33% 0.22% 0% 39.06% 77% .59 / 100%
EMC Corp. EMC 55.17 M 55.40% -21.50% 80% -11.45% 75% 5.74% 2% 1.13 / 100%
Intel Corp. INTC 70.17 M 19.47% -0.49% 0% 2.59% 37% 8.42% 29% .69 / 100%
Microsoft MSFT 88.07 M 17.90% -3.12% 0% 13.65% 78% 21.94% 90% .61 / 80%
Exxon Mobil XOM 24.33 M 16.70% -4.67% 73% -8.30% 83% 2.78% 17% .53 / 85%
Cisco CSCO 57.66 M 13.15% -7.57% 83% -6.66% 79% 1.28% 2% 1.11 / 100%
General Electric GE 34.27 M 10.17% -3.14% 53% -3.83% 75% 2.59% 10% .72 / 93%
Oracle ORCL 36.72 M 7.04% -5.13% 57% -5.66% 67% 7.38% 44% 1.17 / 100%
Applied Materials AMAT 23.23 M 0.59% -2.92% 22% -10.74% 83% -11.13% 48% .41 / 55%
Negative one-year performance

Countrywide Financial Corp. CFC 32.67 M -65.81% -13.93% 58% -24.16% 50% -30.74% 20% .09 / 20%
Merrill Lynch and Co. Inc. MER 24.26 M -36.62% -13.13% 55% -20.67% 82% -23.97% 92% .51 / 83%
Citigroup C 83.32 M -29.26% -14.75% 70% -19.85% 82% -25.81% 96% .61 / 82%
Comcast Corp. CMCSA 31.87 M -26.34% -4.69% 19% -17.69% 96% -19.12% 90% .25 / 35%
Bank of America BAC 28.88 M -15.61% -3.58% 10% -7.83% 71% -9.89% 97% .70 / 62%
Time Warner TWX 21.86 M -10.81% -1.34% 26% -4.47% 70% -3.75% 19% .33 / 95%
Pfizer PFE 37.68 M -10.29% -3.68% 53% -4.38% 71% -2.57% 28% .67 / 95%
Wells Fargo WFC 23.46 M -7.68% -0.32% 0% -1.91% 22% -8.65% 94% .68 / 53%
JP Morgan Chase JPM 23.22 M -5.65% -3.24% 25% -0.13% 0% -3.10% 17% .76 / 90%
Yahoo! YHOO 38.77 M -4.19% -15.34% 67% -2.21% 18% 11.83% 40% .94 / 100%
This information is for educational purposes only. Active Trader provides this data in good faith, but it cannot guarantee its accuracy or timeliness. Active
Trader assumes no responsibility for the use of this information. Active Trader does not recommend buying or selling any market, nor does it solicit orders
to buy or sell any market. There is a high level of risk in trading, especially for traders who use leverage. The reader assumes all responsibility for his or
her actions in the market.

In this section…
INSIDE the Market Morgan must account for
BY JEFF PONCZAK missing e-mails 44
Quick Scalps 44
Exchange consolidation continues 45
Physician, heal thyself More growth for BATS? 46
Hedge-fund group Refco still in news

Managed money
47
47
makes recommendations NYSE eliminates curbs

Stock firms
48
48
NYSE improves block trading 49
Global news 50
W
hile the Securities and pected market events.”
Exchange Com- The recommendations are broken
“MFA has developed
mission (SEC) and down into seven sections: fund manage-
and updated the
the U.S. Congress ment, relation with investors, determina-
have been trying to increase hedge-fund tion of fund value, risk management, reg- recommendations in
regulation for the last couple of years, the ulatory controls, relationship with coun-
hedge-fund industry took a step toward terparties, and crisis management. the belief that the most
policing itself in early November.
The Managed Funds Association Fund management effective form of industry
(MFA) expanded and updated recom- The MFA wants to ensure that a manag-
mendations it made to members in 2005 er’s goal for a fund is consistent with the oversight is self-discipline
after a government group suggested the size and nature of the fund, and that
industry needed more self-regulation. investments and risks are based on the and self-monitoring
The report, which sought input from fund’s specific objectives. Each fund
hedge-fund managers, financial execu- needs to monitor the portfolio managers
by hedge-fund managers
tives, accountants, and investors, focuses it allocates capital to, and should estab-
as part of a shared
on a few major points. lish procedures and policies that deal
Overall, the MFA wants to make sure with changes in software, data, IT, etc. responsibility with market
all fund managers are on the same page And, monitoring third parties providing
when it comes to internal policies and key business functions is critical. counterparties, investors,
practices. The report states, “MFA has
developed and updated the recommenda- Relation with investors and regulators.”
tions in the belief that the most effective Fund managers need to provide ample
form of industry oversight is self-disci- information to prospective investors
pline and self-monitoring by hedge-fund about the fund’s objectives, permissible — The Managed Funds
managers as part of a shared responsibili- investments, and risk factors. Investors Association Report
ty with market counterparties, investors, should be able to evaluate their interest in
and regulators. The recommendations the fund with this information. conflicts of interest. If so, those conflicts
should enhance the ability of hedge-fund Plus, fund managers need to consider need to be disclosed to potential
managers to manage operations, satisfy whether their actions in running the fund investors. Investors need to be notified if
responsibilities to investors, comply with (e.g., relationships with brokers and “soft- a manager makes a side deal or similar
applicable regulations, and address unex- dollar” funding) present any potential arrangement that gives some investors

“There’s a common
belief that if we as fund
preferential treatment.
A fund’s legal counsel is important managers don’t show Relationship with counterparties
here, as it can work with the manager to The fund managers need to make sure
establish what disclosures need to be we’re serious about transactions are executable, and should
made, and when. always seek best execution. Relationships
self-regulation, Congress should be reviewed periodically to ensure
Determination of fund value the fund’s counterparties are still provid-
The report stresses that the assets and lia-
or the SEC will force ing the same services initially sought.
bilities of the fund need to be verified by
an independent auditor. And if the audi-
us into things we Crisis management
tor’s statements don’t match up perfectly The fund manager should have a compre-
don’t want or [things
with the fund manager, procedures need hensive plan for employees and outside
to be established to reconcile the differ- that] are inappropriate.” service providers in the case of unexpect-
ences. ed events such as terrorist attacks, fires,
If the manager is not ultimately natural disasters, etc. This plan should
responsible for making sure the valua- — Hedge-fund manager include a way to protect employees in the
tions are correct, he needs to establish event a crisis makes the workplace
who is accountable and disclose this unsafe, and contingencies for succession
information to the investors. in the fund if a key member passes away.
outputs. When cash needed exceeds Crisis management also includes plan-
Risk management available cash, managers must have a ning for events such as the failure of a
As with fund management, risk manage- plan. This includes keeping an open dia- third-party fund administrator, credit
ment, monitoring, and measurement logue with credit providers. provider, or other important party that
need to be appropriate to a fund’s size would impact a fund’s market, credit, or
and structure. This includes running a Regulatory controls liquidity risk. These plans should be
“stress test” to determine how large Fund managers need to create an envi- reviewed at least once a year, and man-
swings in the market and other risk fac- ronment where compliance with regula- agers should be aware that federal and
tors could affect a fund’s value. However, tors and other authority groups is local authorities can assist in the prepara-
it’s just as important for a manager to stressed. This includes understanding tion of the plan.
realize that any model used to measure or rules and laws in jurisdictions outside “Most of us were doing just about all
manage risk has its limitations and where the fund is located. A code of of these things before, but having the
extreme situations could render the ethics, which would specifically list the MFA compile and distribute them to
model useless. expected behavior of the manager and all managers will hopefully cause the slack-
Before investing in an asset, a manager fund employees, is recommended, as is ers to pick things up,” says a hedge fund
needs to be aware of the liquidity charac- its regular updating. manager. “There’s a common belief that if
teristics of that particular asset. And, as was the case with fund valua- we as fund managers don’t show we’re
A big part of risk management is keep- tion, if the fund manager is not ultimately serious about self-regulation, Congress or
ing track of cash flow, and managers need responsible for making sure all rules and the SEC will force us into things we don’t
to monitor and manage current and regulations are followed, he needs to des- want or [things that] are
future expected instances of inflows or ignate a person for that role. inappropriate.”!

Inside the Market continued
Quick Scalps
The early bird gets the volume Back to arbitration
In an effort to begin trading earlier,
the New York Stock Exchange
announced it would automate the
opening of some stocks while allow-
Deleted e-mails place
ing specialists to handle more active
stocks. Larry Liebowitz, executive vice
president of the NYSE, said the
brokerage at risk
changes would make the exchange
A
more efficient. Although trading
begins at the 9:30 a.m. ET opening
bell, specialists sometimes need a
rbitration has been the mails continued into early 2005.
few minutes if they are responsible
for more than one stock. By allowing savior of some traders Additionally, some of the tapes that stored
some stocks to open electronically, who couldn’t make any the backup e-mails were re-used, erasing
specialists can concentrate on issues headway in a dispute the original content.
where there is an order imbalance.
The exchange hopes the electronic with their brokerage. The brokerage has offered payouts of
opening will persuade more traders However, evidence uncovered in a $3,000 to $5,000 for clients affected by
to send their opening orders to the high-profile lawsuit between Morgan the cover-up, but in many cases investors
NYSE during the first 30 minutes of
Stanley and one of its wealthiest cus- lost hundreds of thousands of dollars.
trading. More than 10 percent of
daily volume occurs during that time tomers could put the venerable Wall The discovery of the deleted e-mails
period. Street firm at risk for hundreds of law- has caused lawyers to prepare hundreds
suits. of new lawsuits against Morgan Stanley.
That toddlin’ exchange Morgan Stanley was sued last year by Darren Blum, an attorney in Coral
The Chicago Stock Exchange set Ron Perelman, a billionaire who made his
records for both share and trade vol-
ume in October. The exchange aver-
fortune by buying out distressed compa-
aged 40.7 million shares daily — a nies. Perelman said Morgan Stanley gave
45-percent increase over September him poor advice on two companies he “We’re not basing our
— and had a 100-percent increase was considering buying, and he lost hun-
from trade-volume records set in
August. Chicago CEO Dave Herron dreds of millions in the transactions. case on the stocks that
says the record volume is an indica- Perelman was awarded $1.45 billion in
tion that traders have recognized the the case, although Morgan won a reversal were affected. We will
attraction of the exchange’s trading
on appeal, and the case is still in legal
model, which is not dependent on
limbo. However, in its defense, Morgan
base our case on how
competing ECNs or slowed by an
order delivery process. As a result of essentially admitted it gave inaccurate
the record volumes, Chicago is
they cheated investors
information, but argued that Perelman
changing its rebate program to give
customers that provide more than 5
was too savvy to fall for it. in arbitration.”
million average daily shares a rebate The information revealed in the trial
of 36 cents per 100 shares. caused the Financial Industry Regulatory — Darren Blum, Attorney
Customers not reaching that thresh- Authority (FINRA) to investigate Morgan
old will receive 32 cents per 100
shares. Stanley, and it found the brokerage did
not provide its clients headed to arbitra- Springs, Fla., has set up a Web site —
How can they be alternative tion access to e-mails that could have www.SueMorganStanley.com — to bring
when trillions are invested in proved their cases. attention to the situation.
them? While Morgan Stanley neither admit-
Morgan initially claimed the e-mails
Alternative assets, which include
were destroyed in the terrorist attacks of ted nor denied destroying the e-mails, the
hedge funds as well as real estate
and private equity, will receive $2.5 Sept. 11, 2001, but FINRA found backup brokerage is expected to sign a waiver
continued on p. 46 files, plus evidence that the deleted e- prohibiting it from taking any action that

Expanding their options
Nasdaq buys Philadelphia exchange
BY JIM KHAROUF AND JEFF PONCZAK
T
he Philadelphia Stock Exchange (PHLX), which has marketed itself as a
would have it directly or indirectly deny potential takeover target for several months, finally got its wish in early
November.
the findings of the FINRA investigation.
The Nasdaq bought PHLX for $652 million in an all-cash deal that will give
Under securities law, the new lawsuits
the Nasdaq instant recognition in the options business. The Nasdaq had
will be settled in arbitration, and Blum
planned to begin its own options market in late 2007, but ownership of the
believes winning the case won’t be much
PHLX gives it an established exchange that has accounted for about 14 percent
more difficult than showing up and pre- of U.S. options volume through the first 10 months of 2007.
senting the settlement. “After an in-depth review of alternatives, we believe that combining with
“We’re not basing our case on the Nasdaq is the best outcome for our customers, shareholders, and the trading
stocks that were affected,” Blum says. community as a whole,” says Sandy Frucher, chairman and CEO of the PHLX.
“We will base our case on how they “No other exchange is better positioned for the future based on technology,
cheated investors in arbitration.” products, and overall passion for continuously redefining the definition and
However, not everybody agrees. value of stock exchanges around the world.”
Chicago-based securities lawyer Dan The board of directors of both exchanges have already approved the deal,
Schramm says investors will have to which is expected to close in the first quarter of 2008, subject to regulatory
approval.
prove their case was damaged by the
The Nasdaq will keep the PHLX’s current structure, which includes a hybrid
missing evidence, and most won’t be
trading platform for options. When the Nasdaq’s options exchange begins oper-
able to.
ations in December, the Nasdaq will be the only one of the six U.S. options
“Not having this information and los-
markets to offer customers a market-maker driven model and a price-time order
ing in arbitration does not automatically book model.
equal fraud,” he says. “Philadelphia has successfully offered floor and electronic trading for some
Additionally, Schramm says, it’s time. We think this capability will continue to be the best approach to serving
possible the deleted e-mails could have options traders as the options market continues to evolve,” says Chris
helped Morgan Stanley’s position, a Concannon, the Nasdaq’s executive vice president of transaction services. “In
defense the brokerage is likely to bring addition to firmly establishing the Nasdaq’s presence in the options market, this
up. acquisition also enhances our organic growth strategy, which will come to
However, the fact that evidence was fruition next month when we launch our price-time priority options platform.”
destroyed will cause the arbitration panel The deal will also give the Nasdaq control of the PHLX’s little-used stock
exchange, as well as a futures business run by the Philadelphia Board of Trade
to question Morgan’s motives, and may
and the Stock Clearing Corporation of Philadelphia.
allow them to overlook the fact that
The deal was made possible only after the PHLX officially settled a class
nothing directly related to the case may
action suit in October.
have been found in the e-mail.
The lawsuit, brought by PHLX shareholder Chuck Ginsburg, was settled in
FINRA would not offer specifics on principle in June but was officially resolved in the Delaware Chancery Court in
the case, but there is some question as to October. Ginsburg challenged PHLX’s 2005 decision to sell 90 percent of the
how valuable the settlement between exchange to six large brokers and Wall Street firms, leaving the original mem-
FINRA and Morgan will be. To win an bers with 10 percent. Ginsburg contended the sale severely — and illegally —
arbitration hearing the second time diluted original shareholders’ stakes.
around, investors would still have to “This is a very exciting time for the exchange and we are delighted that the
prove fraud, and nowhere in the agree- Delaware court has approved settlement of the litigation which has created
ment does Morgan Stanley admit its uncertainty at the exchange for the last 16 months,” Frucher said in a state-
actions were fraudulent, nor does FINRA ment. “This is a good result for the exchange and all of its constituents.”!
make the claim.!

Quick Scalps On the radar
BATS eyes
continued from p. 44
trillion by the year 2011 from investors
two new opportunities

around the world, according to a research
report by Boston-based consulting firm
Casey, Quirk & Associates. Institutional
investment is expected to exceed $1 tril-
B
lion. The report says these assets will be
inexpensive to trade, leading to their
ATS Trading, the upstart viable, competitive, and long-term alter-
appeal, and about one-third of the money
will be invested in swap contracts or other Electronic Communications native to the two incumbent exchanges.”
exchange-listed products that mimic the Network (ECN) that has In October, BATS announced it was
underlying investment. become the No. 3 trading looking into the possibility of launching
venue by volume in less than two years, in Europe a system similar to what it has
Coming up short is trying to increase its presence in the in the U.S.
New York-based Sandell Asset
Management agreed to an $8.2-million U.S. while establishing one in Europe. “We are in the early stages of evaluat-
settlement after the Securities and In early November, BATS filed with the ing setting up an operation in Europe and
Exchange Commission (SEC) found the Securities and Exchange Commission have been approached by a number of
firm illegally sold short the shares of a
(SEC) to become a full-fledged exchange. companies in the past year,” says Randy
New Orleans bank in the wake of
Hurricane Katrina. The SEC said that in “Our motivation to become an Williams, a vice president at BATS.
an effort to offset losses in a long posi- exchange stems from our desire to partic- Beginning in November, European
tion in Hibernia Bank, Sandell shorted ipate directly in the national market sys- markets began trading on the MIFID
shares it did not own, a technique
tem,” says Joe Ratterman, BATS CEO. (markets in financial instruments direc-
known as naked shorting. The firm
gave back the $6.7 million it made from “The BATS organization, and more tive) platform, a system created by the
the trades, plus interest and a importantly our subscribers, will benefit European Union. MIFID is expected to
$650,000 fine. The firm’s founder, in many ways from BATS being directly increase average daily volume at
Thomas E. Sandell, agreed to a
$100,000 fine, and two employees connected to the industry plans. We also European stock markets, and BATS is
were fined $90,000. Sandell and its desire to be on the same regulatory play- hoping to take advantage of that.
employees neither admitted nor denied ing field as our primary competitors — Project Turquoise, a trading platform
any wrongdoing.
Nasdaq and the New York Stock funded and run by seven large European
Exchange.” investment banks, has been in the works
Jury finds Merrill Lynch
full of bull Becoming an exchange would give since last November, but technical prob-
Merrill Lynch was ordered by a Florida BATS an opportunity to list its own lems have delayed its launch, and BATS
jury to pay $6 million to the daughters quotes. Right now, it lists quotes on the could be the beneficiary if it establishes a
of a philanthropist. George Rothman National Stock Exchange and the presence soon enough.
and his wife had $32 million in assets in
International Securities Exchange. MIFID eliminates old rules that
a Merrill Lynch account, but both were
declared mentally incompetent in 1999. Additionally, BATS would gain extra required market participants to use
A jury agreed with the Rothman daugh- revenue by sharing in tape revenue. national exchanges for trading and
ters that Merrill Lynch took advantage of “In terms of the critical metric of reporting, thus opening the door for
the situation by transferring the
Rothmans’ money into investments that matched market share, BATS has clearly alternative systems. NYSE Euronext has
paid higher commissions. George established itself as the third largest mar- already struck a deal with two banks to
Rothman died in 2004. His lawyers say ket center in the U.S.,” Ratterman says. help in trading large orders, while
Merrill Lynch made at least $2.5 million
“Taking the necessary steps to operate our Citigroup and Instinet also have plat-
in fees on investments the Rothmans
knew nothing about. In a statement, market as a registered securities exchange forms in place.
Merrill said, “The verdict is astonishing in is the right thing to do and shows the BATS made its mark in the U.S. by
light of the undisputed fact that the maturity and credibility that BATS brings slashing fees and providing a cheaper
Rothmans, who were wealthy, sophisti-
to participating in the national market alternative to the big two exchanges. It
cated investors, made $10 million on
the annuities at issue, and did not lose system. routinely trades 15 percent of daily
money.” “We want to provide our customers a Nasdaq volume.!

Feeding frenzy
Refco gone but not forgotten
M
ore than two years bankruptcy in the wake of an accounting ed by creditors of Refco, continued its
after Refco, one of the scandal, the attempts to cash in on the litigious ways in October by filing a law-
world’s largest futures situation continue. suit against former Refco insiders seeking
brokerages, declared Refco Litigation Trusts, a group creat- more than $400 million in remuneration.
The “insiders” include former CEO
Phillip Bennett as well as other
Managed futures performance: owners, officers, and directors of
Barclay Trading Group’s September 2007 rankings
Refco. The suit claims the insiders
Top 10 traders ranked by September 2007 return
managing more than $10 million as of 9/30/07.
participated in a massive scheme to
Trading advisor September 2007 YTD $ Under remove assets from Refco, using
return (%) return (%) mgmt. fraudulent transfers within fraudu-
1. Dighton World Wide Inv. (Aggressive) 61.11 38.83 31.5M lent transactions.
2. Dighton World Wide Inv. (SFT 2X) 36.82 37.42 80.1M The Trusts filed a separate lawsuit
3. Clarke Cap'l Mgmt. (Worldwide) 30.57 39.48 17.4M against Thomas Hackl, an executive
4. Fort Orange Capital Mgmt (Gl. Strat.) 30.50 -1.48 12.7M vice president at Refco, for more
5. AIS Futures Management (3X-6X) 28.21 42.71 178.1M
than $5 million. The Trusts contend
Hackl was an active participant in
6. Quality Capital Mgmt. (Gl. Nat. Res) 27.14 2.06 12.0M
the fraud, which led to him earning
7. Quality Capital Mgmt. (Comm. Beta) 26.68 8.67 13.0M
$5 million.
8. Tactical Invest. Mgmt. (Institutional) 26.59 9.78 34.3M
“The lawsuits filed today are in
9. Hawksbill Capital Mgmt. (Gl. Divers.) 25.30 24.88 36.7M
addition to five other lawsuits filed
10. Clarke Cap'l Mgmt. (Gl. Magnum) 24.46 19.19 23.0M by the Trusts and customers of
Top 10 traders ranked by September 2007 return Refco Capital Markets seeking in the
managing less than $10 million as of 9/30/07.
aggregate more than $2 billion dol-
1. Galleon Strategic Mgmt (FX Cannon) 71.59 47.21 1.8M
lars in damages to Refco and its
2. District Capital Mgmt. (Divers.) 41.47 12.88 2.8M
creditors as a direct result of the
3. James H. Jones (Diversified) 37.50 22.56 0.8M
massive fraudulent scheme perpe-
4. Optimus Cap'l Mgmt. (Diversified) 31.68 61.65 0.1M trated for more than eight years by
5. Barbashop LLC 31.66 95.25 4.5M Mr. Bennett, with the aid and assis-
6. Visioneering R. & D. Co. (V-100) 28.52 0.72 1.0M tance of numerous insiders and
7. James River Navigator Fund LLC (3X) 27.09 -11.22 3.8M third parties,” says Marc Kirschner,
8. District Capital Mgmt. (Select) 25.68 3.84 0.2M a lawyer who represents the Trusts.
9. Abundance Fund, LLC 24.07 -14.97 0.7M In the days leading up to the suit
10. TSW Capital Mgmt (Diamond II LP) 20.70 10.71 0.5M of Bennett and other insiders, the
Trusts also brought more than 180
Based on estimates of the composite of all accounts or the fully funded subset method.
Does not reflect the performance of any single account. suits against non-insiders they claim
PAST RESULTS ARE NOT NECESSARILY INDICATIVE OF FUTURE PERFORMANCE. benefited from preferential and ille-
Source: Barclay Hedge (www.barclayhedge.com) gal transfers. Those suits seek more
than $33 million collectively.!

Kicked to the curb
NYSE ends trading halts
W
hen the stock market Some of that drawdown was caused by Stock Exchange established trading
crashed on Black program trades, many of which were “curbs,” which would prevent certain
Monday in 1987, designed to take advantage of arbitrage computer-generated trades from being
the Dow Jones opportunities in S&P 500 stocks. In the made if the NYSE composite index rose
dropped almost 23 percent. aftermath of the crash, the New York continued on p. 49
STOCK PRICES OF TRADING-RELATED FIRMS

Company Symbol Exchange Closing 1-month % 52- 52- Avg. Market
price change change week week daily cap
11/6 high low vol.* ($)
Brokerages
Terra Nova Financial Group TNFG OTC BB 1.65 0.1 6.45% 3.3 1.1 9,510 $44.9M
OptionsXpress OXPS NASD 28.95 1.16 4.17% 32.05 20.78 1.03M $1.82B
Charles Schwab SCHW NASD 23.03 0.2 0.88% 23.61 17.41 6.82M $28.9B
Man Financial Global MF NYSE 30.27 0.26 0.87% 31.61 22 1.17M $3.62B
Track Data TRAC NASD 2.5 -0.001 -0.04% 4.03 2.3 2,800 $21M
TD Ameritrade AMTD NASD 19.18 -0.22 -1.13% 21.31 13.82 3.48M $11.4B
Siebert Financial SIEB NASD 3.5 -0.13 -3.58% 5.76 3.055 17,300 $77.7M
TradeStation TRAD NASD 11.73 -0.5 -4.09% 16.15 9.41 558,000 $521M
AlphaTrade APTD OTC BB 0.22 -0.01 -4.35% 0.29 0.1035 44,700 $9.43M
AB Watley ABWG Pink Sheets 0.041 -0.009 -18.00% 0.14 0.005 125,000 $1.19M
E*Trade ETFC NASD 9.43 -3.92 -29.36% 26.08 8.88 17.1M $4.0B
Exchanges/trading firms
Nasdaq Stock Market NDAQ NASD 48.81 8.31 20.52% 48.39 26.57 3.01M $5.56B
Intercontinental Exchange ICE NYSE 178.34 20.58 13.05% 179.4 81.57 1.89M $12.4B
NYSE Euronext NYX NYSE 90.4 8.85 10.85% 112 64.26 7.19M $23.8B
CME Group CME NYSE 669.1 39.95 6.35% 693 491 676,000 $35.5B
Interactive Brokers IBKR NASD 28.99 1.39 5.04% 34.25 21.00 817,000 $1.17B
International Securities Exchange ISE NYSE 67.09 0.26 0.39% 68.3 39.65 302,000 $2.60B
New York Mercantile Exchange NMX NYSE 134.13 -1.29 -0.95% 152 105 1.17M $12.5B
Nyfix NYFX Pink Sheets 4.35 -0.15 -3.33% 7.5 4.05 56,300 $159M
Penson Worldwide PNSN NASD 16.49 -2.92 -15.04% 34.91 15.44 207,000 $442M
Market makers/specialists
Vander Moolen VDM NYSE 5.14 0.4 8.44% 6.23 4.37 43,900 $234M
Knight Capital Group NITE NASD 13.15 0.54 4.28% 21.78 11.5 1.69M $1.31B
LaBranche LAB NYSE 5.54 -0.13 -2.29% 12.21 4.28 752,000 $341M
Miscellaneous
eSpeed ESPD NASD 10.21 0.93 10.02% 11.28 7.02 142,000 $372M
Interactive Data Corporation IDC NYSE 31.8 1.8 6.00% 32.51 21.73 257,000 $3.0B
MarketAxess MKTX NASD 15.64 -0.7 -4.28% 19.87 11.95 202,000 $483M
Value Line VALU NASD 41.76 -7.99 -16.06% 58.89 40.52 6,350 $417M
*over last three months

Curbs continued from p. 48

or fell more than 2 percent.
However, almost 20 years to the day
after the crash, the NYSE announced it
On the block
would do away with trading curbs, saying NYSE makes a BID(S) for new system
they had lost much of their effectiveness.
B
Curbs have been implemented 18 IDS Trading, an alternative trading system developed by a consor-
times in 2007, but in its filing with the tium of big banks and financial firms, has made its name in the
Securities and Exchange Commission, the marketplace by providing an anonymous venue for trading large
Big Board said the curbs have little impact blocks of shares (more than 10,000).
on volatility. BIDS has been a popular choice for block traders looking for a counter-
“Volatility is neither restrained nor party to make a large trade without giving away their intentions in the
enhanced by the imposition of the col- market. The platform traded more than 1 billion shares (double counted)
lars,” the filing said. “The exchange is in less than 125 trading days.
making this change since it does not Because of BIDS’ success, NYSE Euronext wants in on the fun. In late
appear that the approach of market October, the two groups hooked up in an effort to improve execution
volatility envisioned by the use of these quality and availability of liquidity for NYSE firms trading large blocks.
collars is as meaningful today as when the Both BIDS and NYSE Euronext will have a 50-percent stake in the ven-
rule was formalized in the late 1980s.” ture, which must still gain regulatory approval.
The increase in electronic trading has “We believe this approach with BIDS Holdings will serve our clients and
reduced the impact index arbitrage trades markets well by bringing block-size orders back into contact with active
have on the market. The latest data avail- traders, algorithms, and retail flow,” says Duncan Niederauer, president
able from the NYSE indicates less than 5 and co-COO of NYSE Euronext.
percent of total volume is done through BIDS’ platform allows traders to control how much information about a
those types of trades. trade they want disclosed. They can auto-execute their order or choose to
On Black Monday, program trades negotiate for a better price, they can set a minimum block size, and they
accounted for a little more than 12 per- can choose not to trade with certain parties based on past behavior.
cent of volume, although index arbitrage “We expect that the joint venture will address the current inefficiencies
trades were responsible for only about 3.2 in block trading, such as market fragmentation,” BIDS CEO Tim Mahoney
percent of trading. says. “Moreover, this initiative is an endorsement of our efforts to build the
Even with the curbs eliminated, bro- leading block interest discovery and trading system in the U.S.”
kerages would still have to report any
program trades, which the NYSE has clas-
sified as a basket of at least 15 stocks with
a value of at least $1 million.!

Global News ume is not growing on the same pace

as exchanges such as the NYSE and
Japan’s August jobless rate
increased 0.2 percent from the previous
those in Asia, it is attracting attention month to 3.8 percent, a decline of 0.3
for the first time. percent compared to the same month
EUROPE in 2006.
The London Stock Exchange (LSE)
continues to upgrade its trading sys- Australia’s September jobless rate
The UK’s Q3 GDP rose 0.8 percent tem, announcing in late October it was dropped 0.1 percent from the previous
from the previous quarter and increasing the platform’s speed by 40 month to 4.2 percent, a decline of 0.5
increased 3.3 percent on an annual percent to a trade execution time of percent from the same month a year
basis. The country’s unemployment rate less than 10 milliseconds. TradElect, the earlier.
for June through August remained at LSE’s platform, had an average execu-
5.4 percent compared to the previous tion time of 140 milliseconds when it China’s Q3 GDP grew 11.5 percent
three-month period and fell 0.2 percent started in June, the exchange said. on an annual basis. While the total
compared to the same three months in Capacity was also increased on exceeded China’s forecast of 8 percent
2006. TradElect, from 3,000 orders per sec- and kept the country on track for its
ond to 4,200 orders per second. fifth-straight year of double-digit
Germany’s August unemployment growth, it was down slightly from 11.9
fell 0.1 percent from the previous The European Energy Exchange percent in the second quarter, primarily
month to 6.1 percent, a drop of 2.1 (EEX) and Eurex have partnered to because of export reduction.
percent compared to August 2006. begin trading emissions. Members of
both exchanges will be able to trade The Singapore Exchange’s net prof-
The INSEE, the French government EEX carbon dioxide products through its for the second quarter tripled from
agency responsible for economic releas- the Eurex platform, which will guaran- the previous quarter as volume rose
es in the country, has temporarily tee clearing and settlement of the prod- substantially. Many regional indices hit
stopped publishing its unemploy- ucts. The two exchanges have more all-time or multi-year highs, and the
ment report, saying the usual calcula- than 600 member firms between them, exchange benefited from increased list-
tion methods “were no longer suitable, the largest international network for ings from Chinese firms.
because of the sharply divergent emissions trading. Market makers, who
changes in the two information sources will ensure liquidity, have already been Six traders are suing the Sydney
from which the series were compiled.” selected. Futures Exchange (SFE), claiming
Nonetheless, the INSEE’s October 2007 the SFE cancelled trades that led to
economic analysis report stated “the almost 1 million Australian dollars in
French economy is projected to create ASIA & losses. The trades were cancelled in
nearly 340,000 new jobs in 2007.” The Australian fixed-income products after
group predicted economic growth of AUSTRALIA consumer price index data was
0.7 percent in Q3 and 0.5 percent in released in July. The SFE says the trades
Q4, despite a weaker global economic China’s foreign exchange reserves were erased after someone at the
environment. were more than 1.43 trillion dollars at exchange entered an incorrect number
the end of September, the country’s of trades into the system, causing erro-
Russia’s third-quarter GDP grew 7.4 central bank said. That’s an increase of neous pricing. While the trades made
percent on an annual basis, an increase 45.1 percent from a year earlier. The by the six complainants were cancelled,
from the 6.6-percent growth the coun- country’s trade surplus for the first nine other trades contingent on the original
try enjoyed for the same period in months was $185.7 billion, already deals weren’t, leading to the losses.
2006. Russian economists have forecast more than in all of 2006.
a yearly GDP growth of 6.5 percent, The Dalian Commodity Exchange
although the country is ahead of that Hong Kong’s Q3 unemployment (DCX) in Beijing received approval to
projection. rate fell 0.1 percent from Q2 to 4.1 trade palm oil futures, the fourth new
percent, a drop of 0.6 percent from the product launched by the DCX this year
The Moscow Interbank Currency same quarter in 2006 and the lowest after zinc, rapeseed oil, and linear low
Exchange (MICEX) handled more level in nine years. A government density polyethene. The exchange has
than $2 trillion in trades through the spokesman said sustained economic not specified an exact date but said on
first nine months of 2007, placing it growth offset summer workers return- its Web site it had completed all the
among the top 20 exchanges globally ing to school for the new academic preparations necessary for launch.
for the first time. The MICEX has been year. While China uses more palm oil than
around for 15 years, and while its vol- any other country, it does not produce

any. In 2006, it imported more than 5 cent. Employment gains in educational at 25.5 percent from the previous quar-
million tons of palm oil. services increased by an estimated ter. The rate dropped 0.1 percent from
25,000 and helped to offset declines the second quarter in 2006.
The National Commodity and observed earlier in the summer. This
Derivatives Exchange (NCDEX) of jump spurred public-sector employment The Ethiopian government is hoping a
India plans to begin a spot market — growth to 4.4 percent on the year new commodity exchange will minimize
the NCDEX Spot Exchange — and says through October, while the private-sec- the effects of famine in the country and
bullion will be one of the main prod- tor growth was only 0.4 percent. encourage agricultural growth. The
ucts. The Spot Exchange will allow Ethiopia Commodity Exchange
traders to trade directly, and the prices Brazil’s August unemployment (ECEX) is scheduled to begin trading in
will be set by the international market. rate was unchanged at 9.5 percent December. Its goal is to provide a better
The NCDEX is also awaiting approval to from the previous month and fell 1.1 method of pricing agricultural products
begin trading weather and index percent from the same month in 2006. and minimize risk for producers, hope-
futures. fully leading to more citizens producing
agriculture. Ethiopia’s government has
The Shenzhen Stock Exchange has been notorious for micromanaging the
enlisted help from the Nasdaq as it pre-
AFRICA country’s agriculture, although conven-
pares to launch a Nasdaq-like market tional wisdom is the exchange will help
for start-up firms. The two exchanges South Africa’s second-quarter the country manage surplus quantities
signed a memorandum of understand- unemployment remained unchanged of grains.
ing (MOU) in October, which the
Nasdaq hopes will lead to smaller
Chinese firms listing on the Nasdaq. In
September, the Nasdaq received
approval from the Chinese government
to establish an office in Beijing. The Interest-rate monitor
Nasdaq and the NYSE are the only two
foreign exchanges approved by the
Chinese government, although the UK, ! The People’s Bank of China increased its one-year yuan
Germany, Singapore, Hong Kong, and lending rate 0.17 percent to 7.29 percent in September.
South Korea have shown interest.
After an announcement by the

! The Bank of Norway raised its deposit rate 0.25 percent
to 5 percent in late September. The bank has raised rates 25 basis
Securities and Exchange Board of points 12 times since November 2005.
India (SEBI) roiled the Indian markets
in early October, the SEBI changed its
stance and said it would allow hedge ! The South African Reserve Bank boosted its repurchase
funds greater access to the Indian mar- rate 0.5 percent in October to 10.5 percent. The increase is the
kets. In an effort to minimize foreign third 50-basis-point increase since June and the seventh since June
involvement in the Indian markets, the 2006.
SEBI initially placed restrictions on the
funds, which caused the Indian bench- !
The Central Bank of Turkey dropped its overnight bor-
mark Sensex index to fall 9 percent the
rowing rate 0.5 percent in October to 16.75 percent. The reduc-
morning of the announcement. The
tion is the second in as many months, and it follows a period of
SEBI was initially concerned that a dis-
proportionate amount of foreign invest- inactivity that stretches back to July 2006.
ment could pose a systemic risk to !
India’s financial system. The National Bank of Hungary dropped its two-week
deposit rate 0.25 percent in September to 7.5 percent, the sec-
ond decline in four months after a five-month tightening period in
AMERICAS mid- to late-2006.
Canada’s September jobless rate " The Central Bank of the Philippines dropped its overnight
dropped 0.1 percent from the previous borrowing rate 0.25 percent in October to 5.75 percent. The
month to 5.9 percent, the first time drop comes three months after a huge 1.5-percent reduction,
since November 1974 that Canadian which was the first move in the rate since October 2005.
unemployment was less than 6 per-

THE Economy
U.S. economic briefing
FED CUTS RATES 0.25 PERCENT

Meeting: Federal Open Market Committee
Date and time: Oct. 31 at 2:15 p.m. ET
Summary: After cutting the fed funds rate 0.50
FIGURE 1: QUARTERLY GDP PERFORMANCE Third-quarter GDP climbed 3.9 percent,
percent on Sept. 18, the Federal Open Market according to the advance (first) estimate.
Committee (FOMC) slashed its target rate
Source: Bureau of Economic Analysis Seasonally adjusted *Advanced
another 0.25 percent to 4.50 percent on Oct.
31. The market expected this cut, but some
participants were disappointed it wasn’t bigger.
Although in September the FOMC warned
of a possible economic slowdown, its latest
statement was more explicit. The Fed’s
announcement implied that additional rate
cuts were unlikely in 2007, and also stated
“the pace of economic expansion will likely
slow in the near term, partly reflecting the
intensification of the housing correction.”
The following tables compare the S&P 500’s
daily and weekly reactions to economic releas-
es to its historical behavior since 1997 (or ear-
lier). The market climbed 1.2 percent when
the Fed cut interest rates on Oct. 31, while it
dropped 2.64 percent when the ISM manufac-
turing report was released on Nov. 1.
Historically, however, the market has
climbed the most after Federal Reserve
policy statements and the ISM manufacturing FIGURE 2: PAYROLLS VS. UNEMPLOYMENT RATE Non-farm payrolls increased by 166,000
in October, pushing its three-month average up to 118,000. The unemployment rate
releases.
held steady at 4.7 percent, while its average rose slightly.
Average Source: Bureau of Labor Statistics Seasonally adjusted

Rate S&P 500 historical moves
changes reaction since 1994
Report day 1.20% 0.36%
Five days later 0.70% 0.13%
THIRD-QUARTER GDP
GROWTH STRENGTHENS
Report: Gross domestic product for Q3 2007
(advance estimate)
Date and time: Oct. 31 at 8:30 a.m.
Actual: 3.9 percent
Previous: 3.8 percent
Consensus: 3.1 percent
Average
S&P 500 historical moves FIGURE 3: OVERALL VS. CORE INFLATION Annual gains in the CPI and PPI both spiked at
GDP reaction since 1994 least 0.8 percent in September after falling in August. Core CPI and PPI levels were
essentially unchanged.
Report day 1.20% -0.01%
Source: Bureau of Labor Statistics Not seasonally adjusted
Five days later 0.70% 0.46%

CPI, PPI EDGE HIGHER
Report: Consumer Price Index (CPI)
Actual: 0.3 percent (core 0.2 percent)
Previous: -0.1 percent (core 0.2 percent)
Consensus: 0.2 percent (core 0.2 percent)
Average
S&P 500 historical moves
CPI reaction since 1980
Report day 0.18% 0.07% FIGURE 4: ISM MANUFACTURING INDEX The Institute of Supply Management’s manufac-
Five days later -1.23% 0.17% turing index fell 1.1 points to 50.9 in October — its fourth consecutive decline.
Source: Institute of Supply Management Seasonally adjusted
Report: Producer Price Index (PPI)

Actual: 1.1 percent (core 0.1 percent)
Previous: -1.4 percent (core 0.2 percent)
Consensus: 0.5 percent (core 0.2 percent)
Average
PPI reaction since 1994
Report day 0.48% 0.12%
Five days later -0.92% 0.52%
FIGURE 5: S&P PERFORMANCE

MANUFACTURING GROWTH SLOWS AGAIN The S&P 500 fell sharply on the
day after most economic num-
Report: ISM manufacturing index
bers hit the Street in October
Date and time: Nov. 1 at 10 a.m. and early November.
Actual: 50.9 Source: eSignal
Previous: 52.0
Consensus: 51.5
Average
ISM S&P 500 historical moves
Manufacturing reaction since 1997
Report day -2.64% 0.32%
Five days later -4.76% 0.67%
PAYROLL INCREASE SURPRISES ANALYSTS

Report: Employment
Date and time: Nov. 2 at 8:30 a.m.
Non-farm payrolls
Actual: 166K
Previous: 96K
Consensus: 80K
Unemployment rate
Actual: 4.7 percent
Previous and consensus: 4.7 percent
Average
Employment reaction since 1994
FIGURE 6: S&P 500 REACTION TO NEWS The S&P 500 only closed down once on the
Report day 0.08% 0.13% economic news days in October and early November, despite an overall drop of 5.5
Five days later -2.23% -0.06% percent during this period.

Tech for TRADERS
Software Screening: MultiCharts 2.1

REVIEWED BY DAVID BUKEY
Product summary
Product: MultiCharts 2.1
M
What it is: Chart analysis program with
back-testing features.
ultiCharts 2.1 is an analysis platform with a
flexible programming language that lets you Who it’s for: Stock, futures, and currency traders.
develop custom technical indicators and strate- Skill level: Intermediate to advanced.
gies, test and optimize trading strategies, and Web site: www.tssupport.com
evaluate performance statistics from several perspectives.
At first glance, MultiCharts resembles TradeStation — both Company: TS Support Plus
platforms’ charts and programming languages appear identical. 15 East North Street
Even the name of MultiCharts’ developer — TS Support — Dover, DE 1990
evokes a connection between both products, although none Tel: (888) 817-6385
exists.
E-mail: support@tssupport.com
MultiCharts 2.1 has several unique features. For instance,
it integrates financial data from multiple sources, its charts are Price: 30-day free trial. One-time fee of
intuitive and fun to use, and its strategy-performance reports are $899 or $499 per year. See Web site
detailed and easy to interpret. Also, all indicators, conditional for other pricing options.
studies, and strategies are code-based, which means you can Upside: Object-oriented charts are easy to
quickly change settings or create new trade systems from modify. Programming language is
scratch. compatible with TradeStation’s Easy
The program isn’t perfect. Its help manual is sparse and was Language. Backtesting and
written by Russian software engineers, it lacks a quote window, optimization tools are robust.
and it hogs system resources even though TS Support improved Performance reports are detailed and
performance in version 2.1. But overall, MultiCharts is a reason- easy to navigate. Clever integration
able alternative to other chart-analysis packages that offer similar of multiple data sources.
tools and cost twice as much. But you will need a powerful PC
Downside: No quote window. Help manual lacks
and a tolerance for a few minor bugs.
some details and has been roughly
translated into English. A couple of
Managing data bugs and occasional slow performance.
MultiCharts handles real-time and end-of-day data feeds from
15 different vendors such as Bloomberg, Interactive Brokers, Recommended Windows 2000 or XP,
Patsystems, and opentick — a relatively new source that pro- system 1.2 GHz Pentium 4 or compatible,
vides free and discounted real-time financial data from dozens requirements: 17” or larger dual monitors with
of exchanges. (For a review of opentick, see “Related reading,” 1024x768 resolution, 512 MB RAM
p. 57.) and 1 GB hard drive space, broadband
Few programs use as many real-time data providers or inte- Internet connection, and MS Internet
grate them as effectively as MultiCharts. It can plot the same Explorer 6.0.
stock, futures contract, or currency pair from multiple sources Tested on a 2.66 GHz Pentium 4 with 512 MB RAM
in a chart with different time intervals. If, for example, you were and T1 Internet connection.
tracking Dominion Resources (D), you could plot hourly bars
from Patsystems and five-minute bars from eSignal in a single
chart. You must insert a symbol to MultiChart’s database each time you
Figure 1 shows the QuoteManager and a partial list of data plot a new instrument — an added step.
sources (upper and lower windows, respectively). The You can also import and export price data in ASCII format,
QuoteManager groups all symbols by type and exchange and which is convenient if you analyze prices in Excel. Exporting
lets you customize session times (i.e., electronic vs. pit). We data into a text file was fairly easy, but we had trouble finding
used eSignal’s real-time data feed and had one minor complaint: files after MultiCharts imported them. Clearly we missed a step,

but the manual lacked any information about this
problem.
Charting
Charting is one of the program’s strengths. You can plot
any instrument in any time interval, from ticks and sec-
onds to minutes, hours, days, and so on. For instance,
you can plot 15-tick bars in one chart and seven-minute
bars in another.
Figure 2 shows four charts that contain different sym-
bols in various time intervals. The charts include techni-
cal indicators and strategies that appear on the symbols
or below them.
The platform organizes its charts in standard, yet effi-
cient ways. You can add dozens of charts, customize their
appearance (colors, types, time frame, etc.), and then
save multiple charts as a workspace, which resembles an
individual sheet within an Excel file. Figure 2 shows a
four-chart workspace, but you can create additional ones
and change workspaces with one click. Also, you can
link charts together so that they update when you switch
symbols.
Charts contain innovative features that few other prod-
ucts offer. For example, Figure 2’s Mini Dow chart (upper
left) plots price bars of 1,000 contracts each, instead of
FIGURE 1: MANAGING DATA MultiCharts can manage about 15 data as a function of time. And charts can plot price in certain
feeds, including Bloomberg, Patsystems, and opentick. point intervals (i.e., 10-, 25-, and 100-point bars).
Source: MultiCharts Moreover, Figure 1’s euro futures chart (lower left) shows
hourly bars with 10-minute bars
below them — a setup we haven’t
seen before.
You can access all of
MultiCharts’ important tools
within its point-and-click charts.
It’s easy to add indicators, condi-
tional studies, and strategy signals
to a chart. You can even trade via
Interactive Brokers from this
window. Figure 2’s Mini Dow
chart includes signals from a long
stochastic strategy, and the euro
futures chart (lower left) shows
signals from a strategy that buys
or sells the euro at certain times
of the day.
Figure 2’s five-minute chart of
the Russell 1000 index exchange-
traded fund (IWB; upper right)
includes three conditional stud-
ies: price above its 14-bar aver-
FIGURE 2: CHARTS Chart windows are quite versatile, so you can plot price and volume age, price below that average,
data in countless ways. You can also apply any type of indicator, conditional study, or
and breakouts above 20-bar
strategy to charts.
highs (dark and light blue bars,
Source: MultiCharts
continued on p. 56

Tech for Traders continued
x- and y-axes, drawing trendlines and Fibonacci tools,

and switching chart types (bars, candlesticks, lines).
Finally, removing an indicator is easier than adding one
— a welcome feature.
Indicators and studies

MultiCharts includes 310 technical indicators and condi-
tional studies to plot on charts or add to more complex
trading systems. They are fully customizable and can act
as starting points for further analysis. The indicators range
from the usual suspects — Bollinger Bands, stochastics,
and average directional index (ADX) — to more obscure
ones such as chaos fractals and standard error bands.
Conditional studies identify dozens of price patterns,
including nine candlestick formations (doji, evening star,
FIGURE 3: CODE FOR INDICATORS AND STRATEGIES MultiCharts’ tools are hanging man), and traditional ones such as head and
based on a programming language that is similar to shoulders and pennants. Most studies aren’t original and
TradeStation’s EasyLanguage. This strategy buys and sells were written for TradeStation, but they are quite helpful if
eurocurrency futures at different times of the day.
you want to track objective patterns such as gaps, inside
Source: MultiCharts
and outside days, and breakouts.
Programming language
MultiCharts’ key features are
based on its programming lan-
guage, which closely resembles
TradeStation’s EasyLanguage. This
means you can import any indica-
tor, conditional study, or strategy
code directly from TradeStation
— one of MultiCharts’ selling
points.
MultiCharts claims its code is
95-percent compatible with
TradeStation’s EasyLanguage, but
how it differs is unclear; the help
manual doesn’t describe the intri-
cacies of its code. Instead, the
manual posts links to tutorials on
TradeStation’s Web site.
However, we imported several
TradeStation strategies without
any problems. Figure 3 shows
MultiCharts’ PowerLanguage (PL)
Editor with the code behind the
euro time-zone strategy from
Figure 2 (lower left).
FIGURE 4: PERFORMANCE REPORTS These detailed reports simplify strategy evaluation You don’t have to write code to
and let you break down test statistics in three main categories: equity curve and modify parameters for an indica-
drawdown, individual trades, and different time periods.
tor, study, or signal, but program-
Source: MultiCharts
mers can access this editor from
any chart and revise these formu-
and yellow dots, respectively). las or create new ones. And as you write code, MultiCharts sug-
Like most of its peers, MultiCharts also lets you modify gests different functions that could be appropriate.
charts by selecting the amount of data to plot, resizing the

Back-testing and evaluating strategies
MultiCharts can test multiple strategies against one sym-
bol’s historical price data, but it can’t back test one (or
more) strategies against a portfolio of stocks, futures, or
currency pairs. Despite this drawback, MultiCharts’
back-testing engine is fast and easy to use.
The platform includes 107 canned systems based on
signals such as moving average crossovers, price channel
breakouts, and key reversal bars. Again, you can tweak
the pre-defined systems or write new ones. After you
apply a strategy’s signals to a chart, MultiCharts tests it
automatically against any historical data loaded in that
chart. The amount of historical data you can test FIGURE 5: 3D-OPTIMIZATION CHART After MultiCharts optimizes a sys-
depends on your data feed. In short, if you can plot it, tem’s variables, you can select its ideal parameters with this 3D
you can test it. chart. This strategy gained the most ground with a lower stop-
You can analyze back-test results with a series of loss and a higher profit target (yellow dots, upper left).
detailed reports that break down its performance statis- Source: MultiCharts
tics in three main categories: equity curve and draw-
down, individual trades, and different time periods. requirements, but 512 MB of RAM wasn’t enough to avoid occa-
These reports stand out, because they are well organized and sional problems.
easy to interpret. However, MultiCharts was relatively stable — it froze only
Figure 4 shows a stochastic countertrend system’s equity when we loaded a great deal of data or plotted it in bars of a
curve and drawdown (right) with a list of available performance specific volume (1,000 contracts) or size (25 points). If you have
graphs (left). There are about 38 different tables and charts that a newer machine with at least 1 GB of RAM, you may not run
measure a strategy’s performance: basic statistics and ratios, a list into trouble. And the program is designed to run faster on PCs
of individual trades, and maximum favorable and adverse excur- with multiple CPUs, although that feature wasn’t tested.
sions (among other metrics).
Performance reports are also easy to navigate — you can Support
switch among them simply by clicking a different title. And MultiCharts’ help manual needs work — it lacks detailed infor-
these statistics can be exported into Excel. mation and certain phrases are lost in translation, although the
Russian developers who wrote it understand English well. TS
Optimization Support is incorporated in Delaware, but its headquarters are in
MultiCharts can optimize a strategy’s variables to find ideal val- Rostov-on-Don, Russia, which explains its off-peak technical
ues that can maximize profits or minimize risk. The program support hours. But the support staff always answered the phone
uses two methods of optimization: Exhaustive search and genet- and was eager to answer any questions.
ic algorithms. The exhaustive-search approach calculates all pos-
sible combinations of values to find the best ones. By contrast, Bottom line
the genetic-algorithm approach is faster and uses evolutionary MultiCharts is a clever program with intuitive charts, a fast
tactics to select the best variables. back-testing engine, and strong reporting tools. At $899 it is
After you specify a range of values for each variable, cheaper than some of its peers. But it still has a few bugs and its
MultiCharts generates a spreadsheet of possible combinations help manual is too thin. Despite these minor complaints,
that you can sort according to net profit, maximum drawdown, MultiCharts offers popular features at an affordable price.!
average trade, and so on. In addition, you can create a 3D chart
that plots two variables as a function of one key performance
statistic. Related reading
Figure 5 shows a 3D-optimization chart of the stochastic
countertrend system shown in Figure 4. The yellow dots show “Product review: opentick data,”
this system performed best with a smaller stop-loss and a larger Active Trader, July 2007.
profit target (upper left). This cheap data feed requires some extra work, but it’s
worth checking out.
Performance
MultiCharts back-testing and optimization engines crunch num- You can purchase and download past articles at
bers quickly, but the platform tends to hog resources and bog www.activetradermag.com/purchase_articles.htm.
down at times. Our test machine met the recommended system

The BUSINESS of Trading
A trader tax primer

The beginning of a new year is a good time to review trader tax rules.
In the first part of a multi-part series, we look at what it takes to qualify
as a trader and some of the benefits that come with it.
BY ROBERT A. GREEN, CPA
T he majority of traders still

don’t take advantage of all the
tax breaks they are entitled to.
Unfortunately, far too many account-
ants still do not know about these breaks
are entitled to this tax season and allow
them to fare even better in the current tax
year.
gross income (AGI), and not deductible
against the nasty alternative minimum tax
(AMT), which Congress is seeking to
repeal.
The average trader saves more than
or the many nuances and pitfalls that $10,000 with trader tax status, and hedge
accompany them. They are selling their funds save significant taxes for their
clients short. investors, too.
The IRS lumps all traders into the You can still claim trader tax status for
“investor” category, and investors get the last tax year and even for the prior
penalized in the tax code with restricted three tax years (if you file amended
“investment expenses,” capital-loss limita- returns). Unlike other aspects of trader
tions ($3,000 per year), wash-sale loss tax benefits, trader tax status does not
deferrals, no retirement plans, and more. need to be elected in advance (such as
To get tax breaks, traders must first IRC 475 MTM).
learn the unpublicized rules for traders,
navigate around the vague and strict Qualifying can be tricky
qualification requirements, make certain However, while trader tax status does not
(tricky) tax elections on time, and execute need to be elected in advance, you do
the strategies properly on tax returns need to qualify for it. Full-time traders
(which also is somewhat difficult). generally qualify. Part-time traders can
Accounting for trading gains and losses also qualify, but it’s more difficult. The
is also the responsibility of the securities bar is raised in the eyes of the IRS —
trader, and this adds another layer of especially if they have losses.
complexity. However, this problem can be A business trader meets most of the
solved with the proper software. following qualifications:
And, there are significantly different • Trades full-time or part-time, all day
tax rules and rates for securities vs. com- every day.
modities/futures vs. forex vs. foreign • Spends more than four hours per
futures vs. other types of instruments, Trader tax status day, every day, with few sporadic
and it’s often hard to tell which financial The first step toward tax savings is quali- lapses.
instrument falls into which category. fying for trader tax status, allowing you to • Makes 300 to 500 (or more)
For these reasons, one taxpayer can use business tax treatment as opposed to round-turn trades per year; however,
think he or she owes significant taxes investor tax treatment. there is no magic number in case
whereas another taxpayer with the same Business treatment gives full ordinary law. Forex and futures trades are
facts and circumstances might be entitled loss deductions (including home-office, not listed line-by-line on the tax
to a large tax refund. That’s way too wide education, start-up expenses, margin return (as securities trades must be),
a gulf. interest, and much more) whereas invest- so the IRS cannot see the number of
This roundup of trader tax rules will ment expenses are very limited, only forex or futures trades.
help traders uncover the tax breaks they allowed in excess of 2 percent of adjusted • Makes mostly day trades or swing

trades, and few positions are held to-market (MTM) accounting. If you it’s a different type of MTM than IRC 475.
for more than a few weeks. qualify for trader tax status, you can Forex traders have ordinary tax-loss
Investment positions are segregated. deduct business expenses for tax savings, insurance by default, so they don’t need
• Has the intention to run a but if you have large trading losses in to elect MTM. Forex traders also have the
business and act accordingly with securities or futures, you will be stuck best of both worlds: They can “internally”
formal record keeping, business with capital-loss treatment (capital-loss elect out of Forex IRC 988 for the lower
plans, and other documentation. carryovers), unless you elected IRC 475 60/40 tax rates of futures, or they can
• Has significant business tools, on time earlier in the tax year. stay in IRC 988 for ordinary loss treat-
business expenses, and a Unlike declaring trader tax status, elect- ment. Beginning forex traders are usually
home-office. ing MTM has a deadline: April 15 of the safer with IRC 988, since they may have
• Has a material account size approxi- current tax year for individuals and losses their first year.
mately $20,000 or more, not the LLCs/partnerships and March 15 for S- The election to opt out of IRC 988 is
$5,000 account sometimes Corps. “New” taxpayers (i.e., those who filed internally on a contemporaneous
associated with forex trading. have just formed an entity) may elect MTM basis.
internally within 75 days of inception.
If trading activity is anything less than MTM nuances and gambles
the guidelines stated above, the IRS may Tax-loss insurance As mentioned earlier, electing MTM if
scrutinize and challenge your qualifica- With MTM, restricted capital losses you have capital-loss carryovers is a gam-
tion for trader tax status. In this case, it’s ($3,000 cumulative annual limit) are con- ble. You don’t want to skip MTM and
wise to consult a trader tax expert. verted to unlimited ordinary losses, then generate more capital-loss carry-
unlocking significant tax breaks immedi- overs. Conversely, you don’t want to elect
Statutory vs. case law ately, rather than being stuck with large MTM and then not be able to utilize your
Currently, the IRS does not provide capital-loss carryovers to subsequent tax capital-loss carryovers against your MTM
“statutory law” with objective tests for years. Unutilized trading losses is the ordinary trading gains. A new entity can
how to qualify for trader tax status. biggest pitfall for traders. help here.
Subjective “case law” applies, and it’s MTM also reports year-end unrealized Existing individuals and LLC/partner-
weak on guidance; case law has not gains and losses. MTM is not applied to ships elect MTM for the current tax year
caught up with the online business trad- “segregated investment positions.” So, by attaching an election statement to their
ing revolution. Our previous guidelines traders can also use buy-and-hold strate- prior-year tax return or an extension due
are based on years of experience. gies to defer capital gains and take advan- April 15. Existing S-Corps must file by
Case law has a two-part test to qualify tage of the lower long-term capital gains March 15. Taxpayers perfect the MTM
for trader tax status: tax rates, which require a 12-month hold- election by filing a Form 3115 (Change
1. “Taxpayers’ trading activity must ing period. of Accounting Method) with their cur-
be substantial, regular, frequent, Securities traders should usually elect rent-year tax return (which is due in the
and continuous.” MTM, unless they already have significant following year).
2. “The taxpayer seeks to catch the capital-loss carryovers. MTM ordinary On Form 3115, include a Section
swings in the daily market trading gains cannot offset capital-loss 481(a) adjustment, which is your unreal-
movements and profit from these carryovers. ized gain or loss on Dec. 31 of the prior
short-term changes rather than Short-term capital gains are taxed like year. Since new taxpayers are adopting
profiting from long-term holding of ordinary income. But short-term capital MTM, they don’t need to file a Form
investments.” losses are limited, whereas MTM ordinary 3115.
Case law is too broad and subject to trading losses are unlimited. Hence, MTM Not following both MTM election steps
varying interpretation. The IRS likes to is free tax-loss insurance for securities properly may result in IRS examination
challenge “sporadic lapses” in trading traders. and MTM removal. So be careful. The IRS
activity. There are several open questions However, futures traders should gener- is very technical here.
as well. Do you need to trade every day, ally not elect MTM in order to retain
or is just managing your trading business lower tax rates (60-percent long-term Still to come
on a daily basis enough to satisfy the IRS? rates, 40-percent short-term rates, regard- Next month, we’ll take a closer look at
How does the IRS treat automated trad- less of how long the position has been business expenses, entities, and tech-
ing or trading forex at night? held) on IRC 1256 contracts. niques for filing the proper trader-tax
All taxpayers may elect to carry back return. Remember, though, it all starts
Restricted capital losses vs. futures trading losses three tax years, but with trader tax status and electing MTM
unrestricted ordinary losses they can only be applied against futures on time.!
The key to turning capital losses into gains. IRC 1256 contracts are also
ordinary losses is electing IRC 475 mark- marked-to-market at year-end, but For information on the author see p. 4.

TRADING Resources
NEW PRODUCTS & SERVICES
CQG has released version 7.5, which includes Pre-Trade reporting, and custom AccuCharts. For more information, visit
Analytics, a suite of four new studies — the DOMTracker, www.fxsol.com.
DOMTracker Oscillator, DOMActivity, and Older Orders Ratio.
These studies track activities in the order book away from the PowerShares Capital Management has expanded its family of
inside market. The new Order Ticker uses a numeric display fixed-income ETFs by listing a 1 to 30-year laddered treasury
fashioned after the classic stock ticker to display actions at the portfolio, as well as insured New York and California municipal
inside market and in the order book. In addition, CQG now bond portfolios, on the American Stock Exchange. The antici-
offers Quantity Triggered Stop Orders, which are sent once the pated ticker symbols and ETF names are: PowerShares 1-30
resting amount of orders in the order queue drops below the Laddered Treasury Portfolio (PLW), PowerShares Insured
trader set threshold. Other enhancements include scanning port- California Municipal Bond Portfolio (PWZ), and PowerShares
folios of spreads for pre-set conditions, simultaneous chart Insured New York Municipal Bond Portfolio (PWZ). For more
scrolling linked to news headlines, instant messaging with other information, visit www.powershares.com.
CQG traders, and live chat with CQG customer support.
Also, CQG has added the Singapore Exchange (SGX) to its MarketDelta has introduced a new indicator called the TPD
direct trading connections. CQG has connected its hosted trad- Index, the first of its kind to take Market Profile concepts and
ing gateways to SGX, giving customers the ability to trade logic and apply them systematically. Named after Dr. Thomas P.
exchange contracts. SGX has been added to CQG’s list of trad- Drinka, a professor who has researched Market Profile for close
able exchanges, which includes CME/CBOT, Eurex, Montreal, to 20 years, the TPD Index has a pane that shows the 20-day
Euronext, NYBOT, ICE, DME, NYMEX/COMEX, and SFE. success rate for the indicator. The TPD Index provides a maxi-
Finally, CQG and Strategy Runner have announced the inte- mum of one trading signal each day. It can be used with intra-
gration of their trading platforms. Strategy Runner’s connection day time frames as well, but all the research was done using
with the CQG API will allow their FCM partners to utilize daily and 30-minute data. TPD Index is currently being offered
CQG’s data and order routing services. In turn, CQG customers for a one-time fee of $997. Visit www.marketdelta.com/newsite/
will have access to Strategy Runner services and algorithmic tpdi.aspx for more information and ordering instructions.
trading solutions using Strategy Runner’s robust server-based
technology. For more information, visit www.cqg.com. ICE Futures, the leading soft commodity exchange and a sub-
sidiary of IntercontinentalExchange, is now offering foreign
Market Probability is now offering SectorCheck, a free trading exchange futures electronically on the ICE trading platform.
tool that reports current status and trend information for stock Listing of the foreign exchange futures contracts will occur in
sectors and industries. Its features include instantaneous display phases. In the initial phase, the following futures contracts will
of the one-day return in each sector, averaged over all member be offered electronically 22 hours a day on the ICE platform:
stocks in the sector. A similar report is available for each indus- British pound/Japanese yen (GBP/JPY), British pound/Swiss franc
try. The tool further reports performance rankings of all sectors (GBP/CHF), British pound/U.S. dollar (GBP/USD), euro/British
and industries so users can easily judge where trading and price pound (EUR/GBP), euro/Japanese yen (EUR/JPY), euro/Swiss
movement are most concentrated. With one click, the user can franc (EUR/CHF), euro/U.S. dollar (EUR/USD), Swiss
call up a bar chart that displays the performance in any sector or franc/Japanese yen (CHF/JPY), U.S. dollar/Japanese yen
industry over the previous 30 days. The tool is available to the (USD/JPY), and U.S. dollar/Swiss franc (USD/CHF). These for-
public through the company’s Web site, www.marketprobabili- eign exchange contracts will begin electronic trading on the ICE
ty.com. platform at 8 p.m. ET on Thursday, Nov. 8, for trade date Nov.
9. For more information, please visit www.theice.com.!
FX Solutions has introduced GTS Pro, a forex trading platform
designed to meet the needs of today’s currency traders. GTS Pro Trading Resources is a forum for industry businesses to announce
new products and upgrades. Listings are adapted from company
offers forex traders a wide range of tools including an improved
press releases and are not endorsements or recommendations from
forex calculator for margin, pip, and premium calculations, 10 the Active Trader Magazine Group. E-mail press releases to
trading screen layouts, customizable workspaces, price alarms editorial@activetradermag.com. Publication is not guaranteed.
for any currency pair when a certain price is reached, exportable

BOOKSHELF
Traders, Guns and Money: — contrarian ripple trading — in this book.
Knowns and Unknowns They boast of a trading record that includes
in the Dazzling World of Derivatives 1,225 profitable, round-trip trades in just more
By Satyajit Das than two years while using this technique. They
Prentice Hall, 2006 don’t claim their method will make anyone rich;
Paperback, 334 pages they instead aim to provide a way for middle-
$29.99 class traders to earn some extra money.
Das chronicles his 25-year journey through the

world of financial derivatives, interweaving Swing Trading
advice with tales of the business. Das says that (Trade Secrets Course Book
constructing a designer woman’s shoe, with its with DVD)
almost inexhaustible list of unknowns, is very By Oliver L. Velez
similar to derivatives. This book attempts to Marketplace Books, 2007
provide an entertaining insight into this some- Paperback, 115 pages
times arcane sector of the markets. $29.95
This is the first in a series of book and DVD sets

Active Value Investing: for investors that expands upon Oliver Velez’s
Making Money in Range-Bound Markets 90-minute lecture on swing trading. The book
By Vitaliy N. Katsenelson includes definitions and self-tests to comple-
Wiley & Sons, 2007 ment topics including timing, Japanese candle-
Hardcover, 282 pages stick charting, buy and sell setups, and the
$49.95 technology used to find lists of stocks.
Although all-time stock market highs (and

some notable lows) have dominated news cov- Electronic and Algorithmic
erage in recent months, Katsenelson says focus- Trading Technology:
ing on long-term, sideways range-bound condi- The Complete Guide
tions is key to profiting from this bumpy ride. By Kendall Kim
“Active value investing” — where traditional Elsevier, 2007
strategies are continually recalibrated to adapt Paperback, 203 pages
to changing market conditions — is $59.95
explained, with charts and a practical applica-
tion section. Kim covers three areas of trade
automation — electronic, program, and algo-
rithmic trading — for an audience of financial
Contrarian Ripple Trading: service professionals, institutional investors,
A Low-Risk Strategy to Profiting broker-dealers, and software vendors. Economic
from Short-Term Stock Trades and regulatory issues are discussed, as well as
By Aidan J. McNamara and Martha A. Brozyna statistics and other details behind the trade
Wiley & Sons, 2008 process of execution, confirmation, and recon-
Hardcover, 190 pages ciliation.
$49.95
McNamara and Brozyna are a married couple

that describe their short-term trading method

TRADING Calendar January 2008
LEGEND 1 • Markets closed — New Year’s Day
CPI: Consumer price index

2 • November construction spending
ECI: Employment cost index • December ISM
First delivery day (FDD): The first

day on which delivery of a commodity 3 • November factory orders
in fulfillment of a futures contract can
take place.
4 • December unemployment
First notice day (FND): Also known
as first intent day, this is the first day
• December ISM non-manufacturing
on which a clearinghouse can give • LTD: January currency options (CME)
notice to a buyer of a futures contract
that it intends to deliver a commodity
in fulfillment of a futures contract. The
clearinghouse also informs the seller.
5
FOMC: Federal Open Market 6
Committee
GDP: Gross domestic product 7

ISM: Institute for Supply Management
8 • November consumer credit
Last trading day (LTD): The final
day trading can take place in a futures
or options contract. 9
PPI: Producer price index
10 • November wholesale inventories
PMI: Purchasing managers index
11 • November trade balance
Quadruple witching Friday: A day
where equity options, equity futures, • December federal budget
index options, and index futures all
expire.
12
CBOT: Chicago Board of Trade 13
CME: Chicago Mercantile Exchange
14
NYBOT: New York Board of Trade
NYMEX: New York Mercantile 15 • December PPI

Exchange
• November business inventories
• December retail sales
S M T W T F S 16 • December CPI
30 31 1 2 3 4 5 • December production and capacity utilization
6 7 8 9 10 11 12 • LTD: February crude oil options (NYMEX)
13 14 15 16 17 18 19
20 21 22 23 24 25 26
17 • January Philadelphia Fed survey
• December housing starts
27 28 29 30 31 1 2

18 • LTD: All January equity options; January S&P options (CME); Economic Release
January Nasdaq options (CME); January Russell options (CME); release time (ET)
January Dow Jones options (CBOT) GDP 8:30 a.m.
• January University of Michigan consumer sentiment index (prelim) CPI 8:30 a.m.
ECI 8:30 a.m.
19 PPI 8:30 a.m.
Productivity and costs 8:30 a.m.
20
Employment 8:30 a.m.
21 • Markets closed — Martin Luther King Jr. Day Personal income 8:30 a.m.
Business inventories 8:30 a.m.
22 • LTD: February crude oil futures (NYMEX) Durable goods 8:30 a.m.
• December Chicago Fed national activity index Retail sales 8:30 a.m.
Trade balance 8:30 a.m.
23 Housing starts 8:30 a.m.
24 Production
& capacity utilization 9:15 a.m.
25 Leading indicators 10 a.m.

Consumer confidence 10 a.m.
26 Univ. of Michigan
consumer sentiment 10 a.m.
27 Wholesale inventories 10 a.m.
Philadelphia Fed survey 10 a.m.
28 • December new home sales
Existing home sales 10 a.m.
• LTD: February gold options (NYMEX)
Construction spending 10 a.m.
29 • FOMC meeting Chicago PMI report 10 a.m.
• December durable goods ISM report on business 10 a.m.
• January consumer confidence ISM non-manufacturing report

on business 10 a.m.
• LTD: January gold futures (NYMEX)
New home sales 10 a.m.
30 • Q4 GDP (advance) Chicago Fed

• FOMC meeting national activity index 10 a.m.
Factory orders 10 a.m.
31 • December ECI Federal budget 2 p.m.
• December personal income Consumer credit 3 p.m.
• January Chicago PMI report
Note: For expiration dates of additional commodity futures and options, as well as first
notice and first delivery dates, see the calendar in Futures & Options Trader magazine The information on this page is subject to
(www.futuresandoptionstrader.com). As of press time, release dates were not available change. Active Trader is not responsible
for leading indicators, and existing home sales. for the accuracy of calendar dates beyond
Check www.activetradermag.com/calendar.htm for updated information. press time.

Key CONCEPTS
Treasury refresher
Treasury bonds and notes are debt securities issued by the
Average and median: The mean (or average) of a set of values is United States Treasury. They are considered debt instru-
the sum of the values divided by the number of values in the set. ments because by purchasing them you are loaning
If a set consists of 10 numbers, add them and divide by 10 to money to the Treasury department, which then pays you
get the mean. interest (determined by a “coupon rate”) on a semiannual
A statistical weakness of the mean is that it can be distorted basis and returns the principle when the bond or note
by exceptionally large or small values. For example, the mean of matures on the maturity date. T-bonds and T-notes are
1, 2, 3, 4, 5, 6, 7, and 200 is 28.5 (228/8). Take away 200, and called “fixed-income” securities because of the fixed
the mean of the remaining seven numbers is 4, which is much coupon payment an investor receives while holding the
more representative of the numbers in this set than 28.5. bond or note.
The median can help gauge how representative a mean really T-notes are issued in maturities of two, three, five, and
is. The median of a data set is its middle value (when the set has 10 years; T-bonds have maturities greater than 10 years.
an odd number of elements) or the mean of the middle two The minimum bond or note size is $1,000. For example, if
elements (when the set has an even number of elements). The you purchased a $1,000 10-year T-note with a 4-percent
median is less susceptible than the mean to distortion from coupon, you would receive $20 every six months, totaling
extreme, non-representative values. The median of 1, 2, 3, 4, 5, $40 per year; the $1,000 would be paid back to you on
6, 7, and 200 is 4.5 ((4+5)/2), which is much more in line with the maturity date 10 years from now. A bond or note’s yield
the majority of numbers in the set. is its coupon payment divided by the price — in this case,
4 percent ($40/$1,000).
Treasury futures prices indicate a percentage of “par”
The K-ratio: The K-ratio is used to evaluate the equity curve of a
price, which for any Treasury bond or note is 100. T-bond
trading system (see Lars Kestner’s book, Quantitative Trading
prices consist of the “handle” (e.g., 100) and 32nds of 100.
Strategies, McGraw-Hill, 2003).
For example, 98-14 is a price that translates to 98-
K-ratio is calculated in the following way: Compute the equity
14/32nds or $984.38 for a $1,000 T-bond. T-notes are
curve (e), which represents the value of a trading account used priced in a similar fashion, except they can include one-half
to trade the series (or portfolio). If profits are reinvested, of a 32nd — for example, 98-14+ is 98-14.5/32nds, or
the logarithm of the equity at each bar should be used. 984.53 for a $1,000 T-note.!
Compute the linear regression of the equity curve. The linear
regression fits a straight line to the points in the equity
curve using the least squares method. The slope of the linear valuable for comparing trading systems applied over different
regression (sl) is the numerator of the K-ratio. sets of data, particularly when the number of data points differs.
The standard error (se) of the linear regression measures the
amount by which the equity curve differs from the regression Welles Wilder’s smoothing technique is a type of exponential moving
line. The standard error is defined as the standard deviation of average that uses the following formula:
the error — that is, of the bar-by-bar differences between the
equity curve and the regression line. Smoother equity curves c*(1/n) + ((n-1)/n)*p
have lower se values.
The K-ratio is the quotient of the slope (sl) divided by where
the standard error (se), modified by two terms that attempt
to normalize the indicator. c = current period’s closing price;
The Excel formula to compute K-ratio as it is implemented p = the previous period’s Wilder’s exponential
in AmiBroker (www.amibroker) is: moving average value;
n = the number of periods in the average.
SLOPE(B1:B20,A1:A20)*SQRT(DEVSQ(A1:A20))/STEYX(B1:B
20,A1:A20)/20 For example, if today’s close was 25 and the previous
10-period Wilder’s exponential moving average value was 26,
where the bar numbers are in the A column and the equity the new Wilder’s exponential moving average value would be:
values are in the B column.
25*(1/10) + ((10-1)/10*26) = 2.5 + 23.4 = 25.9
The K-ratio is an excellent tool for comparing alternatives
generated by optimization — that is, it is best used to pick Wilder used this formula in many of the indicators he
which set of arguments are best for a given set of data. It is less described in his book, New Trends in Technical Analysis.

UPCOMING Events ADVERTISING Index
Advertiser
Event: TradeTech Foreign Exchange Asia 2007
Date: Dec. 3-4 Active Trader Bookstore
Location: Conrad Hotel, Hong Kong www.invest-store.com/activetradermag/bookstore
For more information: www.ttfxasia.com
CME Group
www.cmegroup.com/tradeforex
Event: TradeStation’s Futures Symposium
Date: Dec. 6-8 CMS Forex
Location: Hallandale, Fla. www.cmsfx.com
For more information: www.tradestation.com/strategy
DME
Event: Opportunity Finance Network 2007 Annual www.dubaimerc.com
Conference
eSignal
Date: Dec. 11-14
www.esignal.com
Location: Miami, Fla.
For more information: www.opportunityfinance.net Fidelity
www.fidelity.com
Event: MTA Mid-Winter Retreat
FXCM
Date: Jan. 24-26
Location: Don CeSar Beach Resort at St. Pete Beach (outside www.fxcm.com
Tampa, Fla.) Global Forex Trading
For more information: Call (646) 652-3300
www.gftfutures.com
Event: The World Money Show Interactive Brokers

Date: Feb. 6-9 www.interactivebrokers.com
Location: Gaylord Palms Resort and Convention Center,
MB Trading
Kissimmee, Fla.
For more information: www.worldmoneyshow.com www.mbtrading.com
MetaStock
Event: Traders Expo New York
www.metastock.com/at127
Date: Feb. 16-19
Location: Marriott Marquis Hotel, New York PFGBEST.com
For more information: www.tradersexpo.com www.pfgbest.com
ProShares
Event: 24th Annual Risk Management Conference
Date: March 9-11 www.proshares.com
Location: Hyatt Regency Coconut Point Resort and Spa, Terra Nova
Bonita Springs, Fla. www.TerraNovaTrading.com/AT
For more information: www.cboe.com/rmc
Traders Expo
Event: Day Trading Seminar Presented by Joe Ross www.tradersexpo.com
and Rogerio Kirchbaum
TradeStation
Date: March 23-24
www.tradestation.com/9573
Location: Sao Paulo, Brazil
For more information: E-mail info@tradingeducators.com.br Trade The News
www.TradeTheNews.com
Event: Traders Expo Los Angeles
Date: June 23-26 Wealth Lab
Location: Ontario Convention Center www.wealth-lab.com
For more information: www.tradersexpo.com
Xview
www.xview.com

TRADE Diary
Over-eager entry puts trade

at a disadvantage.
TRADE
Date: Tuesday, Oct. 30, 2007.
Entry: Long the iShares MSCI Brazil Index Fund

(EWZ) at 83.87.
Reason(s) for trade/setup: Like many recent

trades, this one is primarily based on capitalizing
in an outperforming market (Brazil) during a tra-
ditionally bullish period (November) — and after
a significant sell-off.
Most major stock indexes topped on or around
Oct. 11, sold off until Oct. 22, then rallied again. Source: TradeStation
But the S&P 500 dropped nearly 2 percent from the Oct. 11
high to a lower high on Oct. 29, while EWZ gained more than
5 percent from high to high, reaching 85.25 on Oct. 29. cult to stomach. However, one must be prepared for such things
As the market has rocketed up the past few days, we will wait when trading a “high-flyer” near its all-time highs.
for a pullback to enter. The market made one more up move on Nov. 6 before turn-
The down day on Oct. 30 provided an opportunity, and the ing back down sharply on Nov. 7, and we decided to get out
trade was filled at 83.87. early. Although EWZ never made it to the initial target, it did
rally to 87.67. Perhaps the initial target was too optimistic?
Initial stop: 79.77, which is 0.22 below the Oct. 26 low. This is Taking some profits above 87 would have helped immensely,
a wide stop, but necessary given EWZ’s current volatility level — even if the rest of the trade got stopped out at a loss.
over the past 10 days the average close-to-close move has been Also, in retrospect we were too eager to enter the trade
1.86 and the average daily range has been 2.53. because we had missed the preceding up move; the “pullback”
on Oct. 30 was miniscule. More patience would have allowed us
Initial target: 88.89, which is intended to be a conservative to enter two or three points lower when a real pullback occurred
point to take partial profits on an expected move to 90, the a few days later.
round-number price looming above the market. It’s interesting when trading markets that, like this one, have
moved into uncharted territory. There are fewer ways to model
price behavior to estimate targets. Extrapolating the Oct. 22-
RESULT Nov. 29 low-to-high move (approximately 13 points) off the
Oct. 30 low would result in an eventual target well above $90,
Exit: 81.99. but this is mostly conjecture. Closes above the 60-day high price
have been followed, on average, by gains of 6.4 percent over the
Profit/loss: -1.88 (-2.3 percent). next 10 days, which would have resulted in a target of 87.85
based on the Oct. 26 high.!
Trade executed according to plan? No.
Note: Initial targets for trades are typically based on things such as the histor-
Outcome: The trade looked brilliant for exactly 24 hours: EWZ ical performance of a price pattern or trading system signal. However, indi-
shot up to 86.80 the day after we entered (Oct. 31). But equity vidual trades are a function of immediate market behavior; initial price tar-
markets sold off around the globe at the outset of November, gets are flexible and are most often used as points at which a portion of the
and although EWZ’s losses were smaller than some, the Nov. 5 trade is liquidated to reduce the position’s open risk. As a result, the initial
drop to 80.54 — while within the initial stop level — was diffi- (pre-trade) reward-risk ratios are conjectural by nature.
Trade Summary
Initial Initial
Date Stock Entry stop target IRR Exit Date P/L LOP LOL Length
10/30/07 EWZ 83.87 79.77 88.89 1.22 81.99 11/7/07 -1.88 (2.3%) 2.93 3.33 6 days
Legend — IRR: initial reward/risk ratio (initial target amount/initial stop amount). LOP: largest open profit (maximum available profit during
lifetime of trade). LOL: largest open loss (maximum potential loss during life of trade).

TRADE Diary
Buy high and sell low?
TRADE
Date: Thursday, Oct. 25, 2007.
Entry: Short the Semiconductor HOLDRS (SMH)

at 33.70.
Reason(s) for trade/setup: This position was

put on to partially hedge a long overall stock
market position.
The broader U.S. market had bounced off the
Oct. 22 low but was still weak. To provide some
protection against another downswing in a vul-
nerable market, we decided to open a short posi-
tion, selecting SMH because of its relative weak-
ness: It had underperformed the broader market
for the past week and on Oct. 25 it was still
moving down — sharply — while the rest of the
Source: TradeStation
market was moving up.
Initial stop: 34.39. However, the stop will be lowered to just portfolio it was supposed to be protecting. Oct. 30 was pretty
below breakeven as soon as the trade is more than 0.50 in the much the final straw: The broader market — and the portfolio
money. — lost ground while SMH made a higher high, higher low, and
higher close.
Initial target: 32.10, which is a little above SMH’s 2007 low and Losing money on both the short and long sides prompted us
the bottom of the October 2006-April 2007 trading range (not
to cover the position early on Oct. 31 — which turned out to be
shown).
the high before stocks went on another slide. Thus, we closed
out a hedge — unprofitably — just when it would have been
useful.
RESULT The hedge was arguably put on too early, but the timing was
based on the market as a whole having rebounded after the Oct.
Exit: 34.24. 22 low while SMH had continued to decline. Perhaps the fact
the SMH was so much weaker than the rest of the market lead-
Profit/loss: -0.54 (1.6 percent). ing up to the trade should have been a signal that it might
bounce back? !
Trade executed according to plan? Yes.
Note: Initial targets for trades are typically based on things such as the histor-
Outcome: It might take some additional time to determine if this ical performance of a price pattern or trading system signal. However, indi-
was: 1) a bad idea, 2) a poorly executed idea, or 3) the wrong vidual trades are a function of immediate market behavior; initial price tar-
instrument at the wrong time. gets are flexible and are most often used as points at which a portion of the
SMH not only failed to continue to exhibit relative weakness, trade is liquidated to reduce the position’s open risk. As a result, the initial
over the course of the trade, it actually was stronger than the (pre-trade) reward-risk ratios are conjectural by nature.
Trade Summary
Initial Initial
Date Stock Entry stop target IRR Exit Date P/L LOP LOL Length
10/25/07 SMH 33.7 34.39 32.1 2.32 34.24 10/31/07 -0.54 (1.6%) 0.33 1.00 4 days
Legend — IRR: initial reward/risk ratio (initial target amount/initial stop amount). LOP: largest open profit (maximum available profit during
lifetime of trade). LOL: largest open loss (maximum potential loss during life of trade).

At January 2008 Issue

Încărcat de

Informații document

Descriere originală:

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

At January 2008 Issue

Încărcat de

Drepturi de autor:

Formate disponibile

PUTTING A PIVOT-DAY SYSTEM TO THE TEST p.

TRADING THE EQUITY CURVE

$4.95 U.S. / CANADA Printed in the U.S.A. www.activetradermag.com

19 Symmetrical reversal bars

26 Trading System Lab

1 www.activetradermag.com • January 2008 • ACTIVE TRADER

Contact Active Trader:

© 2007, Ethan Pines

Active Trader Interview

Other stories: 67 Short hedge results in worst of both worlds.

2 www.activetradermag.com • January 2008 • ACTIVE TRADER

Bubble, bubble, toil and trouble

F rom Jan. 1, 2000 through Oct.

3 www.activetradermag.com • January 2008 • ACTIVE TRADER

Dr. Howard Bandy has degrees in mathematics, physics, engineering, and

Volume 9, Issue 1 Active Trader is published monthly

The information in Active Trader magazine is intended

4 www.activetradermag.com • January 2008 • ACTIVE TRADER

ACTIVE TRADER • January 2008 • www.activetradermag.com 5

6 www.activetradermag.com • January 2008 • ACTIVE TRADER

Treasuries keep rolling Who moved my cheese?

making recommendations or selecting stocks.” a. 63 c. 6

ACTIVE TRADER • January 2008 • www.activetradermag.com 7

The inside track

8 www.activetradermag.com • January 2008 • ACTIVE TRADER

Methodology Understanding the tables

1. Purchases of $50,000 or more. 3. The prior closing price to its highest

ACTIVE TRADER • January 2008 • www.activetradermag.com 9

98 instances Report day Open to low Open to high

Stocks drop on report day…

10 www.activetradermag.com • January 2008 • ACTIVE TRADER

…and trade sideways in the next

ACTIVE TRADER • January 2008 • www.activetradermag.com 11

12 www.activetradermag.com • January 2008 • ACTIVE TRADER

Designing and testing

A recent article (“Market facts:

Try several different ideas

13 www.activetradermag.com • January 2008 • ACTIVE TRADER

18 percent of the time — a net risk-

ACTIVE TRADER • January 2008 • www.activetradermag.com 14

tion is very strong with an r-squared

Expanding the system

FourDayLow = L<PriorLow; PriorLow = Ref(LLV(L,LLBBars),-1); PriorHigh = Ref(HHV(H,SLBBars),-1);

15 www.activetradermag.com • January 2008 • ACTIVE TRADER

• Enter at: close

ACTIVE TRADER • January 2008 • www.activetradermag.com 16

17 www.activetradermag.com • January 2008 • ACTIVE TRADER

Howard Bandy article:

ACTIVE TRADER • January 2008 • www.activetradermag.com 18

Symmetrical reversal bars

BY ACTIVE TRADER STAFF

A n interesting type of price behavior occurs when a mar-

19 www.activetradermag.com • January 2008 • ACTIVE TRADER

FIGURE 3: SYMMETRICAL REVERSAL UP PATTERNS Price tended to

20 www.activetradermag.com • January 2008 • ACTIVE TRADER

the second was followed by mixed price

ACTIVE TRADER • January 2008 • www.activetradermag.com 21

Bonds and the first rule of trading

T hrow a rock down LaSalle

Bonds and notes

22 www.activetradermag.com • January 2008 • ACTIVE TRADER

profit if rates fall

ACTIVE TRADER • January 2008 • www.activetradermag.com 23