TASC Magazine January 2009

Stocks & Commodities V.
27:01 (98, 96, 97): At The Close by Adrienne Toghraie
AT THE CLOSE
Point (T) On The

Laffer Curve
When is it too soon or too late? How much is too much
or not enough?
S ome of the most important questions that a

trader must address have to do with getting
the timing and the amount just right. There is an art to
finding the right balance in life between too early and
too late, and between too little and too much. For a
trader, finding the right point to enter and exit a trade
is critical to making money and limiting loss. Exercis-
ing the right amount of risk is also critical. Too much
risk brings the possibility of losing too much money to
keep you in the game. Taking on too little risk means
that you will severely limit the amount of money you
can make as a trader. So how do you find that point of
balance?
POINT (T) ON THE LAFFER CURVE

If you have studied economics, you are probably
familiar with the Laffer curve. For everyone else, the
Laffer curve shows the relationship between taxes and tax
CHRISTINE MORRISON
Other examples:
revenue. As tax rates increase from low levels, people work
harder and the government collects more taxes, but after a Employee compensation: Too little and they are not
certain point (T), as tax rates continue to increase, people stop motivated to be productive, have a tendency to steal
working harder and the revenues drop off. So governments and sabotage, and you must spend time and money to
want to find that point on the curve that simultaneously replace them; too much and you cannot afford to stay
maximizes effort and tax receipts. in business.
Note that this point is on a curve. It would be easy if that Exercise: Too little and you gain weight and become
point of optimization were exactly in the middle of a nice physically unfit; but too much and you risk injury and
straight line, but that is not how it works. If you plot tax rates exhaustion.
on one axis and revenues derived on the other, what you get
Hard work: Too little and you have lots of spare time
is a half circle and not a straight line, because the results rise
but a serious lack of means; too much hard work and
and then fall, so the optimum point (T) falls somewhere along
you have no life.
that arc.
Many if not most human endeavors have a point (T) as Savings: A society that spends every penny and
well. For example, successful parenting usually means that saves nothing will eventually fall into debt without
you have found that optimal point between being strict and the credit to borrow when needed, with a crumbling
indulgent. If you give your children too much, they become infrastructure and an inflationary economy; but a
demanding, lazy, unmotivated, and unappreciative. But if society that saves every penny and spends nothing
you go too far in the other direction, you create rebellion, will stifle demand for its own production, will like-
destructive behavior, low self-esteem, and resentment. Nei- wise destroy its infrastructure, and will fall into
ther extreme produces children with whom you want to spend stagnation and a deflationary economy.
time or who will grow into self-sufficient, responsible adults.
FINDING THE GOLDEN MEAN
How do you recognize when you are at the optimal point in
your quest to reach your goals and/or when you are too close
to one of the extremes? In my experience, this is a difficult
by Adrienne Toghraie
Copyright (c) Technical Analysis Inc.
Stocks & Commodities V. 27:01 (98, 96, 97): At The Close by Adrienne Toghraie
AT THE CLOSE
issue for most people to address because they come into a to work at a backbreaking pace in order to succeed. In
situation with a set of issues, values, and preconceived fact, he usually finds that the increase in physical and
perceptions that act as filters. These filters make it difficult emotional energy he derives from cutting back his
for most people to objectively measure the true effects of their hours allows him to increase his focus and, thereby,
actions. For example: his profits.
1 Comfort zone filters: If you ask a trader if the level
of the risk he assumes is the optimal level for his 4 Values filters: Money management, for example,
trading, he will invariably say yes, regardless of his extends beyond the issues of risk for a trader. The way
results. The reason for this is that most people assume in which he manages his money once he earns it will
a specific level of risk, not because it optimizes also determine his level of success in trading. If he
results, but because it falls within the area of their spends his money as fast as he makes it and does not
comfort zone. put any into savings, he is not providing for himself
If, however, a trader is given the needed support and his family for the possibility that he might need
to expand his comfort zone commensurate with his a cushion of safety. Spending at an unreasonably high
trading skills, he often discovers that his trading leaps level puts him under greater pressure in his trading
into a higher level of profit. The reverse is true as and may be the cause of risky and unsuccessful
well. If a trader who assumes too much risk deals with trading. So what can be the cause of this kind of
his need for the psychological and physiological nonoptimal behavior?
stimulation he gets from risky behavior, he soon I have worked with traders who come from a
discovers that his profits, too, increase. family or cultural influence that measures an
individual’s worth by his material displays of suc-
2 Modeling filters: If you ask a parent if he is exercis- cess. Rather than being able to figure out what is
ing sufficient parental controls to keep his teenager appropriate spending and money management, a trader
out of trouble, he will often tell you yes, regardless of raised in this value system may not be able to make a
his child’s behavior. The optimum point for a parent well-reasoned decision about what level of spending
to exercise limits on behavior is likely to be based on is optimal for him and his family. Instead, societal
his or her own upbringing, and he will be modeling on and family pressures will prevail.
what he has experienced himself or on what he has
seen, rather than on what is effective parenting. The EYEBALLING IT
same principle can apply to traders who exercise self- The fastest way to tell if you are not even close to the optimal
discipline in the conduct of their trading. point (T) toward meeting your goals and that your means are
If a trader was raised in an environment of ex- not working for you is by asking yourself a series of ques-
treme self-indulgence on the one hand or extreme tions:
self-denial and self-discipline on the other, he may be 1 How close am I toward achieving my goals?
modeling a pattern of extreme behavior without be- 2 Are there certain areas of my performance or my life
ing able to see its long-term consequences. Asking that are creating disharmony?
him to moderate his extreme behavior in order to
maximize results will likely bring about little or no 3 If I continue to do what I’m doing, will my efforts
change because he is unaware of what he is doing. eventually result in success?
4 If I were to intensify my efforts in doing what I’m
3 Emotional filters: The level of effort a trader ex- doing, will I improve the outcome?
pends in reaching his goals may also be extreme. I 5 Are the people I know who are successful in the area
have worked with traders who spend 20 hours a day, I want to be successful in doing the same things I am
seven days a week, at their research and trading. It is doing to achieve their success?
difficult to get these traders to cut back on this
extreme behavior in order to maximize their results 6 Do I get feedback from the people in my life (which I
because they are often motivated by emotional and generally ignore) that I am doing something in the
psychological reasons. A trader who feels fear or extreme (such as I am working too much, taking too
guilt or inadequacy may be driven to compensate by much risk, getting too upset or stressed, spending too
working to the extreme. His emotional needs act as a much money, drinking or doing anything else too much)?
filter so he is not able to see the correlation between
his effort and his results.
If a trader who is being propelled by unresolved
emotional issues acknowledges these issues and re-
The first step is acknowledging that
solves them, he often discovers that he no longer has point (T) exists for each trader.
Stocks & Commodities V. 27:01 (98, 96, 97): At The Close by Adrienne Toghraie
AT THE CLOSE
In other words, the fastest way to acknowledge that you are
not at point (T) is to look at your success. You must measure PUTTING BALANCE BACK INTO THE SYSTEM
your success on two levels: first, the quantitative level (how One of the key factors in finding point (T) in your life is the
much money are you making?) and second, the qualitative fact that objectivity, when it comes to your own life, may not
level (how much effort are you expending and how happy are be within your reach without enlisting outside help. Traders
you in the process?). who play at the extremes of the continuum most often have
unresolved issues in their lives that need to be addressed.
MODELING LAFFER They may have set goals, developed well-thought-out plans
So if our own perceptions prevent us from seeing what the with contingencies, and worked diligently to put them in
right point is for optimal result, how can we find it? Just like place, only to be sabotaged by their own unconscious moti-
the optimum point (T) on the Laffer curve is derived from vations. This is where a trained professional can bring bal-
plotting data on a graph, finding the optimal point for any set ance to a trader’s life.
of activities required to reach a goal can be found through One of the first things I do with a trader who wants to work
keeping track of results and plotting them. with me is to ask him to fill out a trader evaluation. This self-
Suppose you are interested in finding the optimal level of assessment is designed to uncover strengths as well as distor-
risk you need to be taking in order to create profits from your tions, inconsistencies, and patterns of self-sabotage in a
trading. If you assigned a number value to the level of risk trader’s performance as evidenced by his perceptions, atti-
from one to 10 and then kept careful records of the profits you tudes, goals, values, and outcomes.
gained from your trading and plotted them on a graph (with
risk on the x-axis and profits on the y-axis), you would find ARE YOU AT POINT (T)?
that you created an arc graph that starts toward the bottom of If you have come to the conclusion that you need to find the
the left side of your graph, rises to its highest point, and falls point (T) in any particular area of your life, you may want to
off toward the bottom again at the right side of your graph. find out what forces in your life and in your own psychology
If you find the highest point on the arc, you will find the are pushing you to the extreme position. Is it a question of
point of maximum profits at the minimum riskpoint (T). The your own fears and insecurities or of a need to conform to the
arc may not be a smooth or symmetrical half circle, but that expectations of others? Does the thought of changing your
doesn’t matter. The objective is to find that point of maximi- approach make you feel uncomfortable or threatened? Are
zation of effort versus results. you getting feedback from your trading and from those
The point of this exercise is to demonstrate the fact that around you that you are far from the optimal point (T) in your
point (T), that optimal point, is not immediately obvious to us approach to reaching your goals?
because of our preconceived positions on the subject. Only If so, you can conduct your own test to see where that
through objective measurement can we find it. optimal point would be. Along the way, you can also enlist
Another example would be to objectively measure your the support of people trained to help you reach that point of
profits versus the number of hours you work a week. Again, balance. Whichever means you use to find point (T), the first
the answer may be a surprise to you. In fact, you might find step is in acknowledging the fact that point (T) exists for each
that you need to increase your hours! Or you might find that trader, and it is different for each trader — and once you find
your trader’s coach was right — you will make more money it and are able to reach it, you will have found your own
when you decrease your hours to only two trading hours a personal holy grail.
day. But the results for each trader are different, so you cannot
extrapolate another trader’s experience for your own. Adrienne Toghraie is founder and head of both Trading on
Target and Enriching Life Seminars. She may be reached at
www.tradingontarget.com.
S&C

Stocks & Commodities V. 27:01 (92): Books For Traders by Technical Analysis, Inc.
Beat The Market: In- trendlines, Elliott began his career on Wall Street he began
vest By Knowing wave, and more. Dif- mastering the art of trading. His trading phi-
What Stocks To Buy ferentiating between losophy and strategies, and thoughts on
And What To Sell (167 trading ranges and fundamental vs technical analysis and more
pages, $22.99 hard- trend exhaustion has are included in this book.
cover, 2008, ISBN 978- long plagued traders
0-13-243978-7) by in every market sec- Swing Trading
Charles D. Kirkpatrick tor. In a sea of trend- For Dummies
II, CMT, by Pearson following technical (330 pages,
Education, publishing as FT Press. Over the studies, one set of $24.99 softcover,
past 25 years, Charles Kirkpatrick’s contrarian market-timing indicators stand out 2008, ISBN 978-
stockpicking technique has outperformed the from the rest because it is predictive: DeMark 0-470-29368-3)
Standard & Poor’s 500 frequently. His sys- indicators were designed to anticipate trend by Omar Bassal,
tem is easy to use, and this book teaches you reversals. CFA, published
what you need to put it to work in your by John Wiley &
portfolio. He reveals why an active strategy ETFs For The Long Sons. This
based on relative stock rankings is a good Run (306 pages, friendly guide covers the ins and outs of the
route to profit, and how publicly available $29.95 hardcover, risky but profitable swing trading approach to
information enables you to create rankings 2008, ISBN 978-0- investing. It explains all the basics in plain
that almost guarantee performance. You’ll 470-13894-6) by English. By using both technical and funda-
learn how to maximize returns while reduc- Lawrence Carrel, mental analysis, you’ll see how to use both
ing risk, and systematically protect yourself published by John approaches to identify promising securities
against bad stocks and bad markets. Wiley & Sons. Lately, in strongly trending markets. And you will
exchange-traded learn how to calculate investment returns
China Fireworks: funds (ETFs) have be- and manage your risk.
How To Make Dra- come a portfolio
matic Wealth From staple of the retail, individual investors. They Stock Trader’s Al-
The Fastest-Grow- are flexible, transparent, and have the ability manac 2009 (192
ing Economy In The to track asset classes. They have allowed pages, $39.95
World (275 pages, investors to own a liquid, diversified portfolio hardcover 2008,
$27.95 hardcover, of commodities or currencies for a minimal ISBN: 978-0-470-
2008, ISBN 978-0- investment. The ETF expert author will show 22902-6) by Jef-
470-27677-8) by what they are, how they work, and outline frey A. Hirsch and
Robert Hsu, pub- some simple strategies to incorporate them Yale Hirsch, pub-
lished by John Wiley into your investment endeavors. ETFs can lished by John
& Sons. China is on fire and there are now help investors create a broadly diversified Wiley & Sons.
1.3 billion people who suddenly have the portfolio for the least amount of money. While stocks fluc-
ability to create personal wealth. The big tuate, they do so in well-defined, often pre-
investors are scrambling for their piece of The Stock Market Phi- dictable patterns that recur too frequently to
China’s investment pie, but the individual losopher: Insights Of A be the result of random chance. This alma-
investors are being fed misinformation and Soviet-Born, New nac is a must-have investment tool with a
conflicting reports that could lead to disas- York–Bred Hedge Fund wealth of information organized in a calendar
trous investments. This book shows you an Trader (139 pages, format. It alerts readers to little-known mar-
easier, more profitable way to invest in China. $19.95 hardcover, 2008, ket patterns and tendencies that help inves-
Also detailed is how to find industries that are ISBN 978-1-934354-06- tors forecast market trends with accuracy
growing, pick industry leaders, make your 3) by Gennady Favel, and confidence. Professional traders, money
buy, and sell for a profit. published by W&A Publishing. This is not a managers, and journalists rely on the data in
how-to manual for trading, but an examina- the almanac.
DeMark Indicators (193 pages, $29.95 hard- tion of a Wall Street trader’s thoughts on the
cover, 2008, ISBN 978-1-57660-314-7) by stock market, starting with his early attrac- FOR ADDITIONAL INFORMATION
Jason Perl with a foreword by Thomas R. tion to risk-taking endeavors. Starting from www.bloomberg.com/books
DeMark, published by Bloomberg Press. The an interest in market strategy piqued as a www.FTPress.com
DeMark indicators are standalone market- Soviet youth, the author continued his quest www.w-apublishing.com
timing tools, but they can be used with more for quick riches through his college career. www.wiley.com
traditional studies like moving averages, His exploits eventually paid off, and when he
S&C

Stocks & Commodities V. 27:01 (61): Explore Your Options by Tom Gentile
Explore Your Options

Got a question about options? Tom Gentile is the chief options strategist at
Optionetics (www.optionetics.com), an education and publishing firm dedicated
to teaching investors how to minimize their risk while maximizing profits using
options. To submit a question, post it to our website at http://Message-
Boards.Traders.com. Answers will be posted there, and selected questions will
appear in a future issue of S&C.
PUTS DON’T MAKE MONEY greater chance that stock B will move Tom Gentile of Optionetics
WHEN STOCK FALLS above $50 and the call option will be in-
I am trying to understand why I did not the-money at expiration.
make any money on a put option. I am While high levels of volatility make have delta negative when you buy a put
hoping you can shed some light. On options more expensive, so does time. option, you still have vega and theta
October 23, 2008, I bought an XYZ put The more time left until expiration, the risk. The stock fell as you anticipated
option with the 15 strike price. I bought more expensive the options. For ex- and delta worked in your favor. How-
two contracts at $2.10, or $420 total. ample, if stock A trades for $35 and has ever, how much value was lost due to
The stock dropped to near the strike options listed in January and March, the time decay? Looking at theta will give
price. I sold on November 11 for $1.95 March contract with the 50-strike will you a clue. What about volatility? Did
and collected $390. I lost $30 even be more valuable than the January with volatility (that is, the volatility priced
though the stock dropped the way I the 50-strike because the March con- into the option contract, known as im-
thought it would. How can this happen? tract allows two additional months for plied volatility) fall? Vega will tell you
In order to answer your question, it is the stock to move in-the-money. how much you risk from changes in
important that you understand that sev- Understanding the “greeks,” which volatility. My guess is that the loss was
eral factors will influence the value of a are measures derived from an option due to a combination of both: time de-
stock option, not just the stock price. pricing model, can help you make bet- cay and a decline in (implied) volatility.
Certainly, the price of the stock is often ter sense of what is happening with The loss from those two factors was
the most important. If I buy a call and your options. Many option-related greater than the gain from delta.
the stock moves higher, chances are the websites and brokerage firms offer the
call will increase in value. On the other ability to see and compute the greeks. VIX OPTIONS
hand, if I buy a put and the stock moves Some offer both option quotes and I have been looking at trading the VIX to
lower, the put will probably increase in greeks in real-time. take advantage of the present market,
value. Most traders understand this and, Delta, for example, tells us how much so I pulled up the option chains. I see
for that reason, many investors buy puts the value of an option contract will that there are plenty of October and
and calls as a leveraged or cheaper way change for each $1.00 change in a stock December strikes listed, but I don’t see
to play moves in the stock. price. Since puts increase in value when any open interest. Do I have a datafeed
While the stock price is the most a stock price falls, put options have problem, or if I don’t, what would cause
important determinant of an option price, negative deltas. Call option deltas are those chains to not have any open inter-
it isn’t the only one. Changes in divi- positive. If I have a call with a delta of est? It would seem to me pretty risky to
dends and interest rates also affect op- 0.30, it will increase in value by 30 be the only one with open interest if the
tion prices. More important, changes in cents for every $1.00 move higher in the data is correct.
volatility and time also have important stock price. I think the problem is that it is still a
implications for option prices. Vega captures the impact of changes relatively new product and needs time
A stock with very high volatility will in volatility and theta measures time to develop. It also trades off of the VIX
have move expensive option premiums. decay. Both puts and calls have positive futures, and most people know futures
Why? Consider two stocks: Stock A vegas because, as we have seen, in- are not nearly as popular as stocks, and
trades in a 52-week range between $30 creases in volatility make an option neither are the options that trade on
and $40 per share and stock B trades in contract more valuable. Falling vola- them. However, it’s getting more and
a 52-week range between $10 and $90. tility has a negative impact on option more popular as more and more inves-
If both stocks are trading at $35 a share, premiums. Meanwhile, time decay is a tors learn and better understand options
which one has a greater chance of mov- negative for puts and calls. That’s why on this index. The fact that the market
ing to $50 over the next six months? some people call options wasting as- has been so volatile and VIX has been
Answer: Stock B. It is more volatile and sets. Both puts and calls have negative reaching new all-time highs is probably
more likely to reach that level. Conse- thetas. helping to drive the increased activity in
quently, a call option with a strike price Now, to answer your original ques- the contract as well!
of 50 on stock B will be worth more than tion. The reason the put option lost
the same option on stock A. There is a value is due to the fact that, while you S&C

Stocks & Commodities V. 27:01 (12-13): Forex Focus: The RSI Miracle by Hadi Seyedinajad
FOREX FOCUS
Access to foreign exchange trading has opened up exciting trading options for the
retail trader. You can now trade alongside corporations and institutions in a highly
liquid market that is global, traded around the clock, and highly leveraged. Before
jumping into this market, however, we must understand the factors that affect the
forex market. With that in mind, STOCKS & COMMODITIES has introduced Forex
Focus to better prepare the retail trader to participate in the currency market.
THE RSI MIRACLE

If we had to choose just one indicator
to tradethe financial markets, it would EURUSD, M5 1.5040 1.5040 1.5033 1.5033 1.5595
be J. Welles Wilder’s RSI. It does ev- 1.5585
erything you expect from an indicator. 1.5575
It’s an overbought/oversold alert sys- 1.5565
tem, a convergence/divergence detec- 1.5555
tor, and a momentum indicator, but it 1.5545
can do even more. 1.5535
1.5525
T
1.5515
he old saying amid radical tech- 100
RSI(21) 42.9976
nicians, that “the price chart
METAQUOTES SOFTWARE CORP.

70
has everything you need,” is not just a 50
slogan. The information that we can 30
extract from price vs. time data is amaz-
0
ing and I am not sure what this reflects, 10:10 11:30 12:50 14:10 15:30 16:50 18:10 19:30 20:50 22:10 23:45 01:05 02:25 03:50 05:10
human intelligence or the harmony 1 Aug 2008 3 Aug 2008 4 Aug 2008
existing in the universe. By reading a

FIGURE 1: THE RSI MIRACLE. Here you see a choppy market in EUR/USD, while in RSI you see a clear behavior.
price chart we must develop mathemati-
cal functions of price. These are tools to help us interpret the market momentum through the RSI.
raw price data or extract hidden information such as momen- In its simplest form, the RSI gives you an exit signal in an
tum or market pace. optimized moment. It will signal a change in the market phase
There are different ways to measure momentum, each — enter from an impulsive phase to a corrective mode or
differing in their definition and calculation. The one that you change direction — quickly. So having a clear RSI trendline
have learned to interpret and that fills a gap in your market brings you enough confidence to stay with the trend and exit
analysis would be the one you pick for your analysis. I have at the best time.
selected the relative strength index (RSI), which, in addition Usually, when the RSI breaks the current trendline, it is a
to identifying overbought/oversold areas, detecting diver- signal to exit your position. Before entering the market, you
gences, and identifying momentum, also linearizes a trend. need to do a thorough analysis and once you are in the correct
Price tops or bottoms are not always in a relatively straight side of the market — that is, you have captured a price move
trendline. By looking at the RSI, you’ll see that a perfect — you often have a clear RSI trendline, which when broken
trendline often manifests itself. In this article I will present will lead you to exit.
examples using the currency markets. Here’s an example. On the five-minute chart of the GBP/
On the five-minute chart of the EUR/USD in Figure 1, you JPY in Figure 2, at point A, I enter a sell position. The price
can see that although it is a choppy market, the RSI is not so. at point A was breaking the lower trendline of a diagonal
It has moved along its clear trendline. Seeing these lines looks triangle at the end of a five-wave up, confirmed with breaking
like a discovery. This market behavior manifests itself well in a previous RSI up trendline (line 1) in a relatively overbought
RSI movements. area of the indicator. You can see that at point A, RSI has
My next step was to apply this behavior to my trading. So I rebounded from a newly formed trendline moving in the
combined my knowledge of the market using techniques such as opposite direction (line 2) that conforms to the direction of
Dow theory and Elliott wave counts together with reading my sell position. This acts as a confirming tool. Now I just sit
back and wait till the RSI violates the downward trendline.
by S.M. Hadi Seyedinajad At point B, while a relatively deep correction appears to be
Stocks & Commodities V. 27:01 (12-13): Forex Focus: The RSI Miracle by Hadi Seyedinajad
GBPJPY, M5 196.42 196.46 196.32 196.46
197.40
197.15
A 196.90
B
196.65
196.46
unfolding, I am still in the market be- 196.40
cause the trendline of the RSI is still 196.15
C 195.90
trending down. Eventually at point C,
195.65
the RSI breaks above line 2, at which D
195.40
point I exit the position with a profit of 195.15
approximately 90 points. RSI(21) 57.5406 100
One thing to note here is that the RSI 70
(2)
cautioned us before breaking line 2. A (1)
50
new trendline started forming, line 3. (3) 30
Consequently, you might have exited 0
00:05 01:25 02:45 04:05 05:25 06:45 08:05 09:25 10:45 12:05 13:25 14:45 16:05 17:25 18:45
at point D when the RSI was bouncing 24 Sep 2008
along line 3, indicating a change in
trend was approaching. In addition, the FIGURE 2: A COLLECTION OF PERFECT RSI BEHAVIORS, IN GBP/JPY. The RSI has formed a countertrend
indicator was oversold and had not line before breaking line 1, after which it moved along line 2. Again, before breaking line 2, a new line (3) had
formed in the oversold area, alerting the trader for an exit. If you were looking at just the price chart you would
broken below line 3. not have seen these indications.
This is a classic example in how to
read RSI by way of its trendlines. I have
214.40
used it as a confirming tool to make my GBPJPY, H4 195.18 195.62 195.10 195.48
213.35
entry decisions. But often it cannot be 212.25
used by itself. You need to do an overall 211.15
market analysis. 210.05
208.95
The last point I am going to mention
207.85
is how you could stay in a trade, even if 206.75
the trendline is broken. Because of di- 205.65
vergences that occur during the final 204.55
steps of a trend, for example, between 203.45
202.35
the top of a wave 3 and wave 5 (Elliott RSI(21) 53.6270 100
wave counts) of an impulse wave, we 70
can capture the last price swing, even 50
though the RSI may be turning. Wilder wait exit
30
calls these patterns failure swings. You 0
14 May 19 May 21 May 26 May 29 May 2 Jun 5 Jun 9 Jun 12 Jun 16 Jun 19 Jun 24 Jun 26 Jun 1 Jul 3 Jul 8 Jul
do need to exercise some patience, pro- 2008 04:00 20:00 08:00 00:00 12:00 04:00 16:00 08:00 20:00 12:00 00:00 16:00 04:00 20:00 08:00
vided you have enough expertise to
identify a failure swing properly. You FIGURE 3: FAILURE SWING IN RSI. After the RSI breaks the upward-sloping trendline, you could continue
holding your position until it exceeds the last swing support level.
must wait until the last peak (support or
resistance) in the RSI breaks (similar in
concept to the neckline in head & shoulders price pattern). price can capture positions with 100% accuracy. Indicators,
You can see an example in Figure 3. with all of their powerful functionalities, are simply studies
Wilder and analyst John J. Murphy have both gone into that you can use in addition to price chart analysis.
detail about this. However, I don’t feel comfortable applying
this method, since I prefer to exit earlier rather than experi- Hadi Seyedinejad, MSc, is a student of mathematics in Sharif
ence such complex scenarios in RSI. Note that remaining in University of Technology. One of his main mathematical
the market in this situation would be profitable, but not by the application interests is finance and the financial markets.
means of RSI, when it has broken its trendline. You could
surely use the Elliott wave, for example, to let the profits run. SUGGESTED READING
You can also read more about divergences in articles by Murphy, John J. [1999]. Technical Analysis Of The Financial
technicians Barbara Star and Martin J. Pring. Markets, New York Institute of Finance.
Pring, Martin J. [1997]. “Reverse Divergences And
THE TIP OF THE ICEBERG Momentum,” Technical Analysis of STOCK & COMMODITIES,
This is just a glimpse of what you can do with the RSI. Volume 15: December.
Although the examples cited here are for the currency markets, Star, Barbara [1996]. “Hidden Divergence,” Technical
the methodology can be applied to any security. Different Analysis of STOCK & COMMODITIES, Volume 14: July.
situations will arise, and it is best to gain some experience Wilder, J. Welles [1978]. New Concepts In Technical Trading
and familiarity with the indicator. I must mention, however, Systems, Trend Research.
that neither the RSI nor any other mathematical function of S&C

Stocks & Commodities V. 27:01 (72): Futures For You by Carley Garner
FUTURES FOR YOU

INSIDE THE FUTURES WORLD
Want to find out how the futures markets really work? DeCarley Trading senior analyst and
broker Carley Garner answers your questions about today’s futures markets. To submit a
question, post your question at http://Message-Boards.Traders.com. Answers will be posted
there, and selected questions will appear in a future issue of S&C. Visit Garner at
www.DeCarleyTrading.com. Her latest book, Commodity Options, is now on sale.
Carley Garner
FULL SERVICE OR SELF SERVICE? online futures trading. For those who
Should I use a full-service futures and feel they are ready to be on their own, I rity date for each of these securities with
option broker or trade for myself online? suggest using a broker or firm offering various coupon payments, and so on.
This is a decision that is often made reasonable rates combined with effi- A Treasury futures contract is essen-
with little research or thought but has the cient and reliable service. tially a proxy or could even be consid-
potential to have the biggest impact on While low-transaction costs should ered an index for a variety of issues
the bottom line. I would argue that trad- be a priority for those capable of self- within a range specified by the exchange.
ers should put nearly as much effort into directed online trading, it shouldn’t be The securities represented in the price of
finding a broker and determining the the sole deciding factor. There are firms the futures contract includes all of those
service type that fits their needs and offering insanely cheap commission eligible for delivery based on CBOT stan-
desires as they do researching the mar- rates, but you will get what you pay for. dards. Here is where many people get
kets that they will be trading. Should there be a time you have techni- lost — the CBOT has developed a con-
Making a mistake in the choice of cal support problems or customer ser- version factor system to allow the fu-
your service level can be costly. An vice issues, it may be difficult to come to tures contract price to reflect the range
inexperienced trader opting for a deep- a timely resolution. Imagine being placed of available fixed-income issues.
discount online rate may accumulate far on hold for long periods of time to re- Conversion factors are assigned to
more in market losses and trade place- solve whatever issues you may have. each cash market bond or note that meets
ment errors than paying a full-service Then imagine this occurring when you time maturity specifications for the par-
broker would have ever cost. This is not have open positions in a market going ticular futures contract expiration. Each
to imply that a full-service broker will wildly against you. bond or note eligible for delivery will be
assure accurate speculation and prevent Thus, it is critical you balance price provided a conversion factor by the
losses, but using a good broker may and functionality rather than basing your CBOT. The factor allotted by the ex-
avoid some of the common newbie mis- decision exclusively on transaction costs. change represents the price in terms of
takes. In addition, it is possible that When it comes to the costs of trading, percentage, at which $1 par of a security
having an experienced broker to bounce transaction costs are only the beginning. would trade if it had a 6% yield to
ideas off of may improve results. maturity. Bond or note issues with a
Here’s an example. I have witnessed TREASURY FUTURES CONTRACTS coupon less than 6% will have conver-
too many novice traders enter into the We all know what bonds and notes are, sion factors less than 1 to reflect that the
softs futures market through the illiquid but what does a Treasury futures con- security is priced at a discount. Those
NYMEX versions of the contracts as op- tract represent? with coupons greater than 6% will have
posed to the more liquid ICE futures. The Similar to a commodity-based futures conversion factors greater than 1 to reflect
lack of trading volume can lead to tre- contract, Treasury futures represent the that the coupon is priced at a premium.
mendous bid/ask spreads and make it delivery of the underlying asset. Futures As you can tell, the intricacies of the
difficult to exit the market, let alone contracts allow the buyer and the seller conversion factor is beyond our scope
come out with a profit. A trader who to agree upon a price today for delivery here. However, it is important to realize
may have saved $20 in commissions and payment of the asset on a future that the net effect of the conversion
could suffer several hundred dollars (or date. What makes Treasury futures more factor is to link the different prices of the
even thousands) in market losses by not complicated than most is the fact that the bonds and notes eligible for delivery to
knowing enough about futures trading. underlying is more difficult to define create a single price of the correspond-
On the other hand, there are those than it is for, say, corn or soybeans. ing standardized futures contract.
who are familiar with the mechanics of In its simplest explanation, the under- The price of the futures contract typi-
the futures and option markets and can lying instrument for the Chicago Board cally doesn’t correspond to the prices
competently place trades. These traders of Trade’s 30-year bond, 10-year note seen in the cash market. However, it is
probably don’t need the help of a full- and five-year note is a $100,000 US important to note that the current yield
service broker and shouldn’t overpay Treasury security. So which of the spe- will always be the same in reference to
for services. That said, don’t assume cific fixed-income securities is reflected the futures market vs. the cash market.
that some experience in trading equities by the price of the futures contract?
will be sufficient for you to jump into There is certainly more than one matu- S&C
Stocks & Commodities V. 27:01 (28-35): Identifying And Timing With The Special K, Part 2 by Martin J. Pring
INDICATORS
Major Trend Reversals And Short-Term Moves
Identifying And Timing

With The Special K
Part 2
In part 1, you were introduced to the Special K indicator and are not apparent from the summed cyclicality of the Special K.
its application on the weekly and daily charts. In part 2, we will Let’s see how the Special K can be used to identify major
show you how Special K can be used to identify major trend trend reversals and to time pro trend short-term moves.
reversals and for timing pro trend short-term moves.
IDENTIFYING MAJOR TREND REVERSALS EARLY
by Martin J. Pring
The following techniques will help better time primary trend
I mentioned in part 1, my favorite method of reversals:
As displaying the Special K uses a calculation (see

sidebar) based on daily data and exclusively
incorporating simple moving averages like that
1 Observing trendline breaks, such as that in late 2006 in
Figure 1.
displayed in Figure 1. As you can see from this chart, the 2 Identifying a reversal in a series of rising or falling
movements in the daily KST and the Special K are similar. The peaks and troughs. For example, the end of the bull
daily KST shows overbought and oversold conditions, which market in November 2007 was signaled this way.
1600
1550
1500
1450
1400
3 Observing the crossovers
S&P Composite 1350
of the Special K’s smooth- 1300
ing. In this case, I typically 1250
1235.8
use a 100-day smoothing 1200
of a 100-day simple mov- 1150
ing average, as shown in 1100
Figure 1. The series of de-
Special K
clining peaks and troughs 250
Lower peaks and
was still in force in mid- troughs signaled
200
150
September 2008 as this ar-
100
ticle was written. Note that 50
the 100/100 smoothing is 0
not offered as the perfect Daily KST -50

-100
solution but more as a 50
smoothing that appears to
0
work fairly consistently. In
-50
Figure 1, the green and red
arrows show two such S O N D 2005 M A M J J A S O N D 2005 M A M J J A S O N D 2007 M A M J J A S O N D 2008 M A M J J A S
crossovers. FIGURE 1: IDENTIFYING MAJOR TREND REVERSALS EARLY. Here in the S&P 500 Composite you see that the
movements in the daily KST and Special K are similar. The daily KST shows overbought and oversold conditions, which are
USING THE SPECIAL K TO not apparent from the summed cyclicality of the Special K.
TIME PRO TREND SHORT-
TERM TRADES 255
CRB Composite 250
The Special K can also be used in con- 245
240
junction with the daily short-term KST. 235
230
The latter can be interpreted as any other 225
220
smoothed oscillator with overbought/ 215
210
oversold readings, moving average cross- 205
overs, divergences, and so forth. 200
195
The Special K, on the other hand, is 190
185
more of a primary trend swing indicator, 180
as it usually experiences rising peaks
Special K 200
and troughs during a primary bull mar- 150
ket and falling ones in a bear trend. As a 100

50
result, the Special K is normally plotted 0
with three or more years of data so that -50
the most recent primary trend cycles are Daily KST

KST is overbought A B -100
-150
placed in perspective. 50
If we also plot the regular KST under-
0
neath, we derive two benefits. First, we
-50
can use the KST’s gyrations to help iden-
M A M J J A S O N D 2001 M A M J J A S O N D 2002 M A M J J A S O N D 2003 M A
tify short-term reversals in the Special K.
If the KST is overbought and reversing as FIGURE 2: TIMING PRO TREND SHORT-TERM TRADES. Here in the CRB Composite you can see the
Special K and the KST plotted. The Special K is a primary trend swing indicator that identifies short-term
in May 2000 (Figure 2), it is more likely reversals. The KST acts as a confirming tool.
to result in an imminent reversal in the
trajectory of the Special K. gences develop when the KST declines but the dominant
When the KST reverses to the upside and there is little or no longer-term cycles used in the Special K calculation propel it
Special K response, chances are that the bearish intermediate- upward. A good example developed in November 2002 and
and long-term forces are dominating. Under such circum- January 2003, as can be seen by the green arrows.
stances, the Special K will most likely register a new low, The idea of rising peaks and troughs for the Special K is
thereby confirming that the main trend remains bearish. especially important. For example, look at the purple arrows
Several examples of this phenomenon are flagged by the at points A and B. For the KST, that shows a lower low in
red arrows in Figure 2. The most glaring discrepancy devel- November 2002, and for the Special K, a higher trough — if
oped in the August–September 2001 period when the KST you can call it that — at B.
experienced a gentle rally, but the Special K continued in its Finally, there is another way in which short-term move-
decline with no sign of strength whatsoever. Bullish diver- ments can help decide whether a specific short-term KST buy
SPECIAL K AND MA FORMULA FOR METASTOCK WEEKLY SPECIAL K FORMULA FOR WEEKLY CHARTS
p1:=Input(“Enter MA Value”,1,500,52); ROC Time EMA of
Weight Total
p2:=Input(“Enter MA Value”,1,500,26); Span ROC
4 4 x 1 4
(Mov(ROC(C,4,%),4, E)*1)+(Mov(ROC(C,5,%),5, E) 5 5 x 2 10
*2)+(Mov(ROC(C,6,%),6, E)*3)+(Mov(ROC(C,8,%),6, E)*5)+ 6 6 x 3 18
(Mov(ROC(C,10,%),10,E)*1)+((Mov(ROC(C,13,%),13,E)*2)+
8 8 x 4 32
10 10 x 1 10
(Mov(ROC(C,15,%),15,E)*3)+(Mov(ROC(C,20,%),20,E)*4)*1)+
13 13 x 2 26
((Mov(ROC(C,39,%),26,E)*1)+(Mov(ROC(C,52,%),26,E)*2)+ 15 15 x 3 45
(Mov(ROC(C,78,%),26,E)*3)+(Mov(ROC(C,104,%),39,E)*4)*1); 20 20 x 4 80
39 26 x 1 26
Mov(Mov((Mov(ROC(C,4,%),4, E)*1)+(Mov(ROC(C,5,%),5, E)*2)+ 52 26 x 2 52
(Mov(ROC(C,6,%),6, E)*3)+(Mov(ROC(C,8,%),6, E)*5)+ 78 26 x 3 78
(Mov(ROC(C,10,%),10,E)*1)+((Mov(ROC(C,13,%),13,E)*2)+ 104 39 x 4 156
(Mov(ROC(C,15,%),15,E)*3)+(Mov(ROC(C,20,%),20,E)*4)*1)+ Special K 537
((Mov(ROC(C,39,%),26,E)*1)+(Mov(ROC(C,52,%),26,E)*2)+ SPECIAL K FORMULA FOR DAILY CHARTS
(Mov(ROC(C,78,%),26,E)*3)+(Mov(ROC(C,104,%),39,E)*4)*1),p1,S),p2,S);
ROC Time SMA of
Weight Total
zero:=0; Span ROC
zero; 10 10 x 1 10
15 10 x 2 20
20 10 x 3 30
SPECIAL K AND MA FORMULA FOR METASTOCK DAILY 30 15 x 4 60
p1:= Input(“Enter First MA Time Span”,1,500,100); 50 50 x 1 50
p2:= Input(“Enter First MA Time Span”,1,500,100); 65 65 x 2 130
(Mov(ROC(C,10,%),10,S)*1)+(Mov(ROC(C,15,%),10,S)*2)+ 75 75 x 3 225
Mov(ROC(C,50,%),50,S)*1+(Mov(ROC(C,65,%),65,S)*2)+ 195 130 x 1 130
(Mov(ROC(C,390,%),130,S)*3)+(Mov(ROC(C,530,%),195,S)*4);
530 195 x 4 780
Special K 2485
Mov(Mov((Mov(ROC(C,10,%),10,S)*1)+(Mov(ROC(C,15,%),10,S)*2)+ SIDEBAR FIGURE 1
(Mov(ROC(C,20,%),10,S)*3)+(Mov(ROC(C,30,%),15,S)*4)+
Mov(ROC(C,50,%),50,S)*1+(Mov(ROC(C,65,%),65,S)*2)+
(Mov(ROC(C,75,%),75,S)*3)+(Mov(ROC(C,100,%),100,S)*4)+ zero:=0;
(Mov(ROC(C,195,%),130,S)*1)+(Mov(ROC(C,265,%),130,S)*2)+ zero;
(Mov(ROC(C,390,%),130,S)*3)+(Mov(ROC(C,530,%),195,S)*4),p1,S),p2,S); zero:=0;
or sell signal is going to work. Note the horizontal line

marking the short-term low in the Special K in May 2001. See how the Special K can be used to
When the indicator violates this level, it signals that a new identify major trend reversals and to
low in the price itself is likely. In the case of the May 2001
example, the Special K took out its low just about two weeks
time pro trend short-term moves.
before the CRB itself did.
The reverse situation developed at point B, where the experience a persistent linear uptrend, individual sectors are
Special K moved to a new high. In this case, there was no lead continually rotating leadership roles. For example, between
by the momentum indicator as the price broke out more or 1998 and 2000, tech stocks performed admirably, but financials
less simultaneously with it. were not so strong. In such situations, indicators such as the
Special K are able to isolate emerging new leadership at a
THE SPECIAL K relatively early stage.
AND RELATIVE ACTION In this respect, Figure 3 features the relative strength line for
On occasion, we see markets experience long-term linear up- the S&P Financials, an early leading sector. When the line
or downtrends; examples would include Japan in the 1980s or is rising, it means that the financials are outperforming the
the US in the 1990s. Such trends do not lend themselves to S&P and vice versa. One of the exchange traded funds
momentum interpretation because oscillators, whose calcula- (ETFs) for this sector is the XLF. You can examine the long-
tion assume a normal four-year business cycle, peak prema- term cyclical swings in the KST in the bottom panel and the
turely and trigger several false sell signals prior to the ultimate Special K in the center. The arrows point to Special K
secular turning point. Because the calculation of the Special K moving average crossovers, which often occur simulta-
assumes the cyclicality of the four-year business cycle, it too neously or close to trendline violations. As you can see, the
suffers from this weakness during linear up- or downtrends. cyclicality of this series consistently offered good signals in
However, the Special K does lend itself to trends in relative this 15-year period.
action. While market averages, such as the S&P, might Figure 4 features the relative action of the S&P Materials,
S&P Financials (XLF) relative to S&P

0.30
see the same thing develop almost a year

0.25
later at point Y.
0.227
0.20 IT’S SPECIAL

Special K Although this new indicator has a lot of
400
300 benefits, it is certainly not the end-all and
200
100 be-all. Each security has its own unique
0
-100
set of market participants, and collec-
-200
-300
tively, they have their own characteris-
-400 tics, as reflected in the price action. Con-
-500 Long-term KST
sequently, with any indicator it is impor-
200 tant to see what kind of characteristics
100
0 have manifested in the past. Are Special
-100
-200
K moving average crossovers consis-
-300 tent? Does the indicator offer extended
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
trends of rising or falling peaks and
FIGURE 3: SPECIAL K AND RELATIVE ACTION. The Special K moving average crossovers often occur at
troughs? Have such reversals offered re-
the same time or close to trendline violations. Note the long-term cyclical swings in the Special K and KST in
the S&P Financials. liable signals?
If you can satisfy yourself that a set of
a sector that does well at the end of the cycle and therefore is specific characteristics work consistently, then go with the
moving (in relative terms) in the opposite direction to the Special K. If not, recognize this weakness and use it with another
Financials most of the time. If you compare Figures 3 and 4, market or security.
you will see that the Financials sector was declining in 2007–
08, whereas Materials was advancing. Martin J. Pring is the president of Pring Research and
The character of the Special K for Materials is also slightly chairman of Pring Turner Capital, a money management
different in that it is subject to strong intermediate swings firm. He is the author of several books, including the classic
within the 2003–05 primary Special K bear market and the Technical Analysis Explained and Introduction to Technical
2005–08 bull market. I have drawn some bullish and bearish Analysis, a workbook/CD-ROM tutorial. He may be reached
trendlines and their respective violations for the relative at Pring.com.
strength line itself. Note how the bearish dominant interme-
diate- and long-term cycles pressured the Special K lower SUGGESTED READING
when the KST was rising at point X. This was negative until Narcouzi, Christopher [2001]. “KST Revisited,” Technical
all three series violated down trendlines in October 2005. We Analysis of STOCKS & COMMODITIES , Volume 19: August.
Pring, Martin J. [2008] “The Special K,”
Technical Analysis of STOCKS & COM-
0.21
MODITIES , Volume 26: December.
0.20
S&P Materials relative to S&P _____ [2004]. “Do Price Patterns Really
0.19
0.18
Work?” Technical Analysis of STOCKS
0.175
0.17
& COMMODITIES, Volume 22: April.
0.16
_____ [1992]. “Identifying Trends With
The KST Indicator,” Technical Analy-
0.15
sis of STOCKS & COMMODITIES Vol-
0.14
ume 10: October.
0.13 _____ [2003]. Introducing The KST, DVD,
0.12
Special K 500
Pring.com.
450 _____ [1992]. “K ST And Relative
400
350 Strength,”Technical Analysis of
300
250 STOCKS & COMMODITIES, Volume 19:
200
150 November.
100
50 _____ [1997]. Martin Pring’s Introduc-
Daily KST
0
-50
tion To Technical Analysis, McGraw-
X Y 50
Hill.
_____ [2002]. Technical Analysis Ex-
0
plained: The Successful Investor’s
-50 Guide To Spotting Investment Trends
S O N D 2003 A M J J A S O N D 2004 A M J J A S O N D 2005 A M J J A S O N D 2006 A M J J A S O N D 2007 A M J J A S O N D 2008 A M J J A S And Turning Points, 4th ed., McGraw-
FIGURE 4: CYCLES IN THE MARKETS. Here you see that the S&P Materials sector is subject to strong Hill.
intermediate swings within bull and bear markets.
S&C

Stocks & Commodities V. 27:01 (54-60): Interview: Forecasting Simplified, With Cynthia Kase by J. Gopalakrishnan
INTERVIEW
Taking Your Trading Seriously
Forecasting Simplified,
With Cynthia Kase
Cynthia Kase, president of Kase and Co., CTA, is considered by many to be the
energy market’s premier technical analyst and hedging advisor. Educated as
an engineer, she worked as a trader and risk manager for Chevron, Chemical
Bank, and the Saudi Oil Ministry’s consulting arm, Petronal, before launching
her company in 1992, which primarily focuses on providing trading and
hedging strategies, software and solutions to the energy market, but also offers
a trading toolbox, StatWare, on a range of trading platforms. She has been the
winner of the Market Technicians Association’s Best of the Best Award, 1997,
received the Key Women in Energy Award twice, first in 2002 for Vision and in
2004 for Innovation and Creativity, and in 2007 became the first American to
be awarded the coveted Master of Technical Financial Analysis diploma from
the International Federation of Technical Analysts.
STOCKS & COMMODITIES Editor Jayanthi Gopalakrishnan interviewed Kase
on November 1, 2008.
C ynthia, tell us how you got

started trading and what you
are doing now.
I was educated as a chemical engi-
trated for a while, because here I was
an engineer but I was in the middle of
this fundamental, gut-feeling kind of
trading instead of numbers and analy- Just learn a market and
neer. I got a master’s degree in chemical sis. When I discovered technical trade it up and down on the
engineering and I worked for the first 10 analysis, I loved it because it gave me
years of my career as an engineer. I something to study and learn and pure technicals. That’s the
graduated in 1973 and worked for helped me get a sense of the market. thing to do.
Polaroid near Boston where I grew up I was transferred to New York in
and then Stauffer Chemical in the Bay 1985 to manage that office’s clean prod- cause in those days, and it still is to some
Area. In January 1980 I went to work for ucts business, including gasoline, jet fuel, degree, technical analysis was associ-
Standard Oil Co. of California, which and heating oil. So I traded physical ated with voodoo. I went to work for
became Chevron. The company had a cargos in New York for the rest of the Chemical Bank as a commodity deriva-
management development program 1980s, but I learned a lot about technical tives trader at a time when the deriva-
where it would move people from engi- trading because trading became, even on tives business was just in its infancy. I
neering to the business side of the oil the physical side, just a differential to the had a lot of time on my hands, so I
industry. In 1983 they transferred me NYMEX. So if you were going to trade a started developing my own indicators
from engineering to the trading group, cargo of heating oil, for example, it was and I bought all the TAG (Technical
and that’s how I became a trader. So NYMEX plus or minus something. It was Analysis Group) tapes that the old Dow
really, it was by accident. a proxy for the physical. Jones Telerate used to offer.
Also, in working for one of the huge
Was that a big change? international oil companies when bro- What else did you do?
As an engineer I would read all the kerage commissions were around $35 a I would also call up people like George
engineering magazines to try to im- round turn in 1985–88 dollars, the bro- Lane and ask questions about
prove my skills and knowledge. When I kers were only too happy to give us lots stochastics. Bill Williams was another
went over to trading in 1983, it was the of education. Back then, the NYMEX person I would call frequently. Then I
same year the crude oil contract was was only a small exchange that nobody went to work for the Saudis for a year
introduced. So there were people trad- had heard of. and got invited by Tim Slater to go on
ing physical but weren’t sure what short this world tour along with Perry
or long meant or what a moving average What did you do after that? Kaufman. I was surprised to find out
was, or anything like that. I was frus- I became a self-taught technician be- that I was regarded as a well-respected
technician because my whole back- And in the early 1990s the natural gas above the 135 level and we are talking
ground in technical analysis was in a market was in its infancy, and crude oil now on November 1, 2008, producers
self-taught microcosm. was very quiet and would trade a five- might be thinking about maybe liqui-
Then I joined the Market Techni- cent range on the day. So it just wasn’t dating that position within the next three
cians Association (MTA) in 1991. Back working. I had to do something else. to six months because it’s not going to
then, very few people in the energy go that much lower. We still think oil
business were involved with technicals. What did you end up doing? will come off another 10% to 20%,
Even though I didn’t think I was doing I struggled about what to do. People which is $6 to $12. So we are getting
anything special, I found myself emerg- suggested that I write a market forecast- fairly close to the bottom. We might
ing from years of teaching myself tech- ing newsletter. I didn’t want to do that. liquidate a position in this case. But for
nical analysis as somebody who was I started to teach technical analysis. the most part they are hedgers and not
just developing new ideas and unique People again suggested that I write a traders, and they have different meth-
techniques. newsletter and do some forecasting. At ods. If they take a position on, they may
Chemical Bank I did a bit of that for the be in and out once over a year. So I have
What then? bank’s clients. It was either that or I these models to help people hedge.
I always wanted to be independent. I would have to get a real job again. So I I have also done a lot of risk manage-
worked for the Saudis for a year and started writing a newsletter in 1993 on ment around the hedging process. This
when my contract was up I couldn’t see natural gas and crude oil and began to includes things like setting trader risk
myself going back into corporate do some risk management. limits, Monte Carlo simulations, and so
America. I didn’t want to conform to the on, which is essentially what I do now.
corporate culture anymore. And you What kind? Over the course of all this time I have
have to remember that I graduated from What I mean by risk management is developed my own forecasting tech-
college in 1973, so the whole politically consulting on hedging. And so now, 16 niques. I’ve been working on a book
correct atmosphere was not something I years later, we developed into primarily and quantifying some of my research. I
was used to, especially since I was used an energy forecasting, trading, and risk have the time to do all that since I have
to being the only woman engineer in the management advisory firm, but we don’t an assistant, Dean Rogers, who man-
middle of a lot of men. It became diffi- do a lot of traditional consulting. We ages my day-to-day operations. He has
cult for me to conform to the idea of write a forecast every week, one on gas learned my forecasting techniques so I
having to watch everything you say. and one on crude oil, and we have a 96% am freed up from my day-to-day chores.
Since I had been a trader for a bank, and accuracy calling the market. I have trad- I also have time to do some extracur-
the only thing the bank cared about was ing software called StatWare carried on a ricular work such as working at a home-
making money without worrying about bunch of trading platforms. That’s how less shelter.
what your ethnic background was or most of the retail end of the business
what your political views were, I found knows me. And within the energy sector You developed some of your own indi-
being a trader very freeing, so I really I have proprietary hedging models. cators. How did those come about?
wanted to be independent. I have two momentum indicators I
What are some differences between a developed, and they have been around
What was stopping you? hedging model and a trading model? for a long time. I won the MTA “Best of
I had been divorced and remarried The big difference between a hedg- the Best Award” for relative strength
and wasn’t in an economic position to ing model and a trading model is simple. and momentum in 1997. I was sur-
trade. You hear stories about people A hedging model places positions 12 to rounded by “quants,” folks with finance
who start with $5,000 and make mil- 18 months at a time. Rather than trading degrees, with high-level math and sta-
lions, but I owned a home and only had the January contract, for example, a tistics skills.
$30,000 in the bank. I tried to trade, but producer may sell forward the whole As a trader I was never really ex-
I was used to trading a cargo of gaso- calendar of 2009 to lock in prices for the posed to statistics and probability theory.
line, for example, which is the equiva- year. So instead of looking at individual I took one course on it in grad school.
lent of 250 contracts, and at the bank we contract months, the producer looks at You can go a long way with a bit of
had a 5,000-contract limit. With $30,000 the whole forward curve at once and statistical knowledge. I sat next to some-
I felt comfortable trading just three con- then most of these positions are held to body who had a doctorate in economics
tract lots and I just couldn’t take small expiration. If you are going to liquidate and I learned a lot from him about the
lot sizes like that seriously. I was used to them, they always get liquidated under kind of math that usually option traders
trading other people’s money, I was narrow conditions of high profitability would know, such as standard deviation
used to trading professionally, and for where the odds of them continuing to be and probability theory. What I did was
me to sit in front of a computer all day profitable are diminishing. use some of this mathematical knowl-
with just three lots was a huge change. For example, if a producer hedged up edge to reengineer the math inside of

momentum indicators. DevStop, which is a statistically based have to look at all these different rela-
indicator that I wrote an article about for tionships. And so we have a technique
How? your magazine in 1991 or 1992, as well where we just define all the waves on a
Most option traders’ assumptions is as a color-coded entry indicator called series of charts and we calculate targets
that prices are normally distributed. If Kase Easy Entry System, or KEES. These based on phi. So we use extensions that
you have a move — say, a three–stan- indicators are embedded in StatWare. some people refer to as Fibonacci ex-
dard deviation move of volatility — tensions but they are 0.62, 1, 1.38, 1.62,
price is relative to volatility, so you can Which means? and a few more.
expect price to mean revert. Volatility is It means that with my indicators 80% If you look at the numbers we use,
the standard deviation of logarithmic of the time you get a signal the market 95% of these patterns meet these phi-
rate of change of price. So to the extent turns, and 80% of the time when the based targets. So we list our waves on
that prices on a log basis have changed market turned it was preceded by a sig- an Excel spreadsheet. Across the top we
relative to volatility, you can measure nal, whereas with the stochastic 80% of have all the different targets and we fill
whether they are trending. the time you get a signal the market turns, in the prices that match those targets in
If you have a high measurement of but only about 40% of the time the mar- the cells. It’s a matter of probability.
price change relative to volatility, you ket turned was it preceded by a signal. Rather than try to figure out the waves
can say prices are trending. And if you and sort out the wave counts and figure
have an extreme move, you can say Can you tell me about your wave out which wave is 1, 2, 3, 4, 5, and which
prices are going to mean revert. If you analysis? is going where, we list all the waves and
look at, say, the logarithm of the high I’ve been using the technique in one all the all the phi targets simply show up
today divided by the low n days ago form or another since 1993 with the in the proper cells.
divided by the volatility over n days, newsletter and before that with the bank.
you know how many standard deviation But I never published anything about it What do you look for in the numbers
prices moved. So that’s the theory the because it was one of the things I used entered in the cells on the spreadsheet?
momentum indicators are based on. for my forecasting so I didn’t want to We use a relatively simple concept.
But what is also interesting is that the tell everybody what I was doing. We look at what numbers come up the
logarithm of the high n days ago is most, which is known as the confluence
variable. What the indicators do is they So you changed your mind? of numbers. Say you have a pattern on a
do a loop from three days back, or 65 I’ve come around in two ways. I’m gold chart. Gold dropped to a low of 774
days back, or however many days back, never going to be able to retire unless I and we filled in all the waves that went
and they look at which time frame was write this book and leave a legacy about down to 774 — from the high to 774 to
the most significant for the trend. So all how I do the analysis. It’s something I all the waves down. We calculate each
this math is buried in the momentum need to do. wave’s targets, and just picked out which
indicators and what is different about The other thing is you can look at any numbers came up the most often; 736
them is that they are what I call bidirec- endeavor. Take a great violinist, for was a number that was highly confluent
tional. They work at about 80% accu- example. Even if he stands on a stage and the market went to 736.4. This is not
racy in both directions. and tells you every little trick of his a complicated technique. It’s just a mat-
technique, it doesn’t mean everybody ter of the more waves you put on the list,
What does that mean? can pick up the violin and play the way the more accurate it is.
If you get a momentum divergence he does. So I am no longer at the point You can forecast using waves with-
on the stochastic, it’s highly reliable in where I have to keep my work a secret out any wave counts. All you have to do
that the market will turn a couple of so I have begun to write the book about is look at what numbers are the most
standard deviations. The problem is of- my techniques. confluent in these grids.
ten the market turns two or three stan- The second tier is if you look at the
dard deviations and you don’t get a In that case, what can you tell us? proportions that exist within the existing
signal. So when you get a signal, it’s It’s pretty simple. Instead of looking pattern. Rather than force it into a five-
great. But then you get a lot of turns with at five-wave Elliott waves, which tend wave count we look at what kind of a
no signals. With my indicators, when to be complex (there are four variables count, a three-wave a-b-c up, makes sense.
you get a signal it’s just as good as the — wave 1 relative to wave 3, wave 1 You can have an abc/ABC or an ABabc/C
stochastic, but you almost always get a ratio to wave 5, wave 3 ratio to wave 5, for example. We label the waves based on
signal when the market is going to turn. and wave 1 to the end of wave 5), I look what the best fit would be to the existing
So it works in both directions and not at three-wave patterns. Fortunately, the proportions. So if you get a five-wave
just in one direction like the stochastic. three-wave pattern only has wave a to pattern where the proportions among the
wave c. It’s very simple. waves are not Fibonacci, we are not going
What other indicators do you have? If you have a, and you have b, you are to label it as a five-wave. If a three-wave
I’ve got another indicator called the just projecting one number. You don’t pattern matches the Fibonacci propor-
tions we will label it a three. So we put the them in the program, calculate all the trading way bigger numbers. You should
labels on it that best fits the existing pro- targets, and spit out which ones are the have at least a couple of years’ living
portions. most confluent. And then we color the expenses in the bank plus enough money
The other thing we do in addition to confluent ones on the grid. to have it be worthwhile for you to start
what numbers come up the most often is trading.
we look at what numbers are generated Do you find it more difficult to call
by the early waves relative to the later these volatile markets? What do you think is the difference
waves. We don’t find it any more difficult to between equity markets and the com-
project the market week to week now modity markets?
What’s the difference? than we ever did. It’s easy to predict The big difference is equity traders
If you have a number generated by what the market will do this week and are stock pickers. They try to pick what
the early waves, it is more important the week after. But some people want is going to outperform the rest of the
because it’s been in the system longer. you to predict where the end of the trend market. It’s more like being single vs.
Then we look at the ways the patterns will be. And that’s always a treacherous getting married. When you are single
lay out in the table. If you’re projecting endeavor, especially since random you try to pick what the best date is for
natural gas, you can get the first wave events can intervene. the week, whereas if you’re married,
projected to maybe $5.50 with the But right now, the markets are de- you stick with the same person through
“smaller than” target for the first wave cently supported where they are in both the ups and downs. So with the com-
(0.62), “equal to” for the second wave natural gas and crude. The challenge is modities market it’s a matter of choos-
(1), the “intermediate,” which is 1.38 whether these support points are going ing the commodity that you can get to
for the third wave, and the “larger than” to hold. One thing about natural gas is know and feel comfortable with. So
target (1.62) for the fourth wave. So you that it is a seasonal commodity. People even though you are trading techni-
can have a series of waves listed, all of always seem to get surprised that it gets cally, you get to know the feel of psy-
which project to the same price, using cold in the winter, even though it does chology when the stats come out — for
extensions mostly based on phi. In this every year. example, storage numbers in the energy
example, there are four waves that So people are optimistic and as soon markets, and how the crowd psychol-
project to $5.50, the first using a 0.62 as it gets cold, people start buying, so ogy is for the hot weather vs. cold.
extension, the second 1.00, the third sometimes we don’t see reality intrude Just learn a market and trade it up and
1.38, and the fourth 1.62. So if you color again in the markets until winter is over. down, up and down on the pure
in the cells, you can see a diagonal technicals. That’s the thing to do.
pattern. How do we know whether the energy
Or you may have a situation where if market is recovering? Any other advice for traders?
you have a wave xyz, there are some A sustained close over $76 in crude I never had to paper-trade because I
calculations in the xy part of the wave and over $7.25 in natural gas. Even was always trading professionally, but
and other calculations on the yz part. If though we are expecting an attempt to it would be a good idea to paper-trade.
a wave cycle is defined by three prices recover, we know it’s not going to be The other advice I give traders is that
xyz, there are instances where a target easy any time soon, and are expecting you have to take it seriously. Always
calculated by the xy part of the wave, $50-ish in crude and around $5.70 on have stops in. That’s very important.
called the impulse portion of the wave, is gas. Also learn, learn, learn — we are start-
equal to a target calculated by the yz, or ing a free educational webinar high-
corrective, portion of the wave, which You said something earlier about not lighted in this issue’s Classifieds.
uses phi, phi-squared, and phi-cubed to starting trading with only $5,000 in
calculate targets. We find that about one- your trading account. Do you have a Thank you, Cynthia.
third of the waves exhibit a pair of xy and threshold amount for what we should
yz calculations that generate the same have in our accounts before we trade?
target. So for example, if there is a 38% The common wisdom is that you
retracement, the “equal to” xy target will have to have $50,000 or $100,000 to SUGGESTED READING
be the same as the phi-squared yz target. lose. That’s reasonable. You have to be Kase, Cynthia [1991]. “Choosing A Time
Using confluence and patterns of tar- in a comfortable-enough economic po- Bar Length In Intraday Trading,”
gets is a way of forecasting that is highly sition so you are trading with the rational Technical Analysis of STOCKS & COM-
accurate and depends more on how many side of your mind instead of the emo- MODITIES, Volume 9: August.
waves you’re patient enough to type tional side. What I found when I was _____ [1993]. “Redefining Volatility
into your spreadsheet than being a smart trading my three contracts was that I had And Position Risk,” Technical Analy-
technician. At this point we’ve auto- trouble taking it seriously. That was be- sis of STOCKS & COMMODITIES,
mated the process to pick the waves, put cause I had been a corporate trader and Volume 11: October.
S&C

Stocks & Commodities V. 27:01 (36-41): Join The Band by Marco Alves
Applying Hysteresis To Moving Averages
Join The Band

Apply this method to moving average crossovers to get rid of when the faster average crosses below the slower one. DAVID GOLDIN
the lag and the false signals. As any technical analyst knows, these crossovers are prone
to whipsaws; the price moves just enough in one direction to
by Marco Alves trigger a signal, then quickly changes direction, triggering an
opposite signal. This causes early entries and exits that
ne of the first indicators that any technical analy- jeopardize trade performance (Figure 1).
O
sis novice studies is the moving average cross- Whipsaws are the result of the sensitivity of MAs to data
over. Moving averages (MAs) smooth a price fluctuations. The classical approach to this problem has been
series by determining the average closing price to increase the averaging period (Figure 2) at a cost of
for a determined period (the last n bars) and, as increased lag, which, if too pronounced, may render the
a result, are lagging indicators, more suited to indicator useless.
trending markets rather than rangebound ones. In addition, whipsaws tend to affect different time frames
If two MAs of different periods are used together, a simple in a similar manner. For example, a set of two MAs on a daily
trading system can be built around it easily. Every time the chart will probably incur a similar frequency of false signals
shorter (faster) moving average crosses above the longer during the course of seven months (154 bars) as an equally
(slower) one, a buy signal is generated; a sell signal is produced parameterized set of MAs will sustain on a three-year weekly
NEW TECHNIQUES $DJAAG (Dow Jones-AIG Agriculture Sub-Index) INDX ©StockCharts.com

31-Dec-2007 Op 80.51 Hi 81.20 Lo 80.31 Cl 81.00 Chg +0.01 (+0.01%) ▲
$DJAAG (Daily) 81.00
MA(20) 79.25 77.5
MA(50) 75.53
chart (3 * 52 = 156 bars). The annoyance 75.0
frequency just moves to a larger time frame. 72.5
So the problem lies in the concept: simple line 70.0
crossovers must be replaced as signal generators 67.5
by a different kind of triggering system.
65.0
If we were to draw charts by hand (the only
option before the computer age), we would 62.5
find the idea of a thick-lined pencil a useful 60.0
one, as it would allow for some breadth where
57.5
the averages intersect. But line thickness is,
STOCKCHARTS.COM
mathematically speaking, an absurdity. What 55.0
we need is a band or an envelope around the
slower MA, within which the indicator can O N D 06 F M A M J J A S O N D 07 F M A M J J A S O N D
jitter around without causing false signals. FIGURE 1: WHIPSAWS. The arrows indicate clear signals, while the circles point to whipsaws. Note
Before we develop this idea any further, let that, for the purpose of clarity, price bars were removed and only the SMA lines were plotted.
us enter the world of engineering and get
acquainted with a crucial notion in control $DJAAG (Dow Jones-AIG Agriculture Sub-Index) INDX ©StockCharts.com
systems. 31-Dec-2007 Op 80.51 Hi 81.20 Lo 80.31 Cl 81.00 Chg +0.01 (+0.01%) ▲
HYSTERESIS MA(40) 76.55 75.0
MA(100) 72.77
First as an engineering student and later work- 72.5
ing in the food industry, I was introduced to
70.0
the area of control systems and to the problem
of repeated activation-deactivation cycles. 67.5
Take, for example, a thermostat controlling a 65.0
cooling system like the one you will find in
your refrigerator. We want it to keep the 62.5
temperature as constant as possible, say five 60.0
degrees Centigrade (5˚C) (T0), and the cool-
ing system can only be in one of two states: off 57.5
or on. If the thermostat reacted immediately to
55.0
any difference from T0, it would activate or
deactivate the system with a frequency that O N D 06 F M A M J J A S O N D 07 F M A M J J A S O N D
would cause stress to the equipment and cool-
FIGURE 2: SLOWING DOWN. Observe how whipsaws were prevented by increasing the periods used
ing inefficiency. by the moving averages, but also note the delay of at least one month in the signals.
To prevent rapid switching on and off as the
temperature drifts around the set point, the
thermostat will turn the system on only when
temperature rises above 6˚C (Ta) and will not
turn it off until the temperature drops below
Temperature
4˚C (Tb). In this situation, we say that the Ta

thermostat has hysteresis.
T0
A system with hysteresis exhibits path-
dependence. When the input value is between Tb
Ta and Tb, we cannot predict the output with-
out looking at the history of the input, the path
that it followed before reaching its current Time
value (Figure 3). In the refrigerator example, if FIGURE 3: PATH-DEPENDENCE (THERMOSTAT EXAMPLE). The colors indicate the state of the
the temperature is 5.5˚C we cannot predict system (output): green when the cooling is on; red when it is off. Above Ta, the system is always on; below
whether the system will be on or off. Tb, it’s always off. Inside the band, the state depends on the input’s path.
Hysteresis importance goes beyond
manmade devices. Many pivotal processes occurring in BACK TO THE TRADING SYSTEM
living cells use hysteresis to help stabilize them against the We are now in a better position to design a signal trigger that
various effects of erratic chemical changes. Similarly, we reduces false signals without increasing the lag. The only
should use this phenomenon to protect our MA crossover desired lag should occur where the lines intersect under
system from localized fluctuations. choppy price moves.
FORMULAS
Function Formula
(period of n bars; date t; 0.1 ≤ k ≤ 0.8; high (H), low (L) and close (C) prices)
t
Simple moving average C + ... + C t ± 1 + C t Σ Ci

SMA t = t ± n + 1 n = i=t±n +1
n
Exponential moving average EMA 1 = C 1
EMA t = EMA t ± 1 + 2 C t ± EMA t ± 1

n+1
C t ± n + 1 + 2 * C t ± n + 2 + ... + n ± 1 * C t ± 1 + n * C t
Weighted moving average WMA t = n
Σ
t
C i ± SMA t 2
Standard deviation SDt = i=t ± n+1
n
Mean absolute deviation

Σ
i = t ± n+1
Ci ± SMA t
MADt =
n
Average true range TR 1 = H 1 ± L 1

TR t = Max H t ± L t ; H t ± C t ± 1 ; L t ± C t ± 1
n
Σ TR i
ATR n = i=1
n
n ± 1 * ATR t ±1 + TR t
ATR t = n
Moving band Moving average ± k * Volatility
Example: Bollinger Band BB(n,k)t = SMAt ± k * SDt
KC n,k,p = EMA t ± k * ATR t

Example: Keltner Channel
n+1
p is the period for ATR. Usually p = 2
If we replace the slower moving average with a moving dispersion. They assess how widely spread a set of values is
band, we effectively substitute one threshold value with two from its mean. The closer the datapoints are to the mean, the
values (above and below the previous threshold) and thus smaller SD (or MAD) will be. ATR was developed by J.
create a buffer against price fluctuations. Buy (sell) signals Welles Wilder to ascertain volatility in commodities prices,
are triggered when the faster MA crosses above (below) the and it uses intraday price variations to compute its value
upper (lower) limit of the band. (see sidebar “Formulas” above).
But how do we set the bandwidth? Fixed amplitudes (a To create our band, we must calculate its upper/lower limits
static difference to/from a threshold) are not adjustable to by adding/subtracting a chosen measure of volatility to/from
longer-term price evolution. They may be too wide for low the threshold (the slower moving average). This always sets
prices or too narrow for high ones. Percentage amplitudes are bandwidth at two times the volatility, and that raises a
too rigid to accommodate changes in price volatility. They problem. Whole measures of SD/ATR produce bands that
may be too ample for calm periods and too tight when prices are too wide for hysteresis purposes (for example, two SDs
swing rapidly. above and below the average will encompass about 95% of
Because we want to reduce the sensitivity to price instabil- any normally distributed price variation). So we should use
ity, we need some kind of measure of price fluctuation and a fraction of those measures in order to build effective
apply it to the band setup. Three such measures are the bandwidths (multiplying volatility by a factor in the range
standard deviation (SD), the mean absolute deviation (MAD), of 0.1 to 0.8 should work in most cases, but experimentation
and the average true range (ATR). with other values may be necessary).
SD and MAD are widely used measures of statistical The sidebar includes formulas for different types of MAs
$DJAAG (Dow Jones-AIG Agriculture Sub-Index) INDX ©StockCharts.com
31-Dec-2007 Op 80.51 Hi 81.20 Lo 80.31 Cl 81.00 Chg +0.01 (+0.01%) ▲
MA(20) 79.25 77.5
BB(50,0.8) 73.09 - 75.73 - 78.37
75.0 It is notable that the MA divides the band in
72.5 two lanes: an upper one, bounded by the MA and
70.0 the higher line, and a lower one between the MA
67.5 and the lower boundary. It is also evident that
during an uptrend in prices — bull mode — the
65.0
faster MA will spend most of the time above the
62.5 slower MA (opposite behavior under a bear
60.0 mode, of course).
With this knowledge, we can adapt our trig-
57.5
gering system to the primary trend in place.
55.0 During a bullish trend, the lower lane provides
a better hysteresis band, since it will provide a
O N D 06 F M A M J J A S O N D 07 F M A M J J A S O N D lower threshold value for buy signals; under a
FIGURE 4: MA CROSSOVERS WITH HYSTERESIS. By replacing one threshold line with a band, we
bearish trend, the upper lane will accommo-
have given the faster MA some space to fluctuate without causing whipsaws while preserving the date the need to sell at higher prices. If the
relevant signals with a minimum lag. trend weakens and prices move sideways, you
should consider returning to the original full-
and volatility indicators. You may use any combination of width hysteresis until a trend is established and you can
moving average/volatility you find helpful. We are trying to select an appropriate lane. Under prolonged periods of range
set up the most useful hysteresis band. We do not want to be trading, MA crossovers become quite ineffective, and other
constrained by unneeded statistical formalities or a pre- indicators such as oscillators would be better suited.
defined indicator. Each person has their preferred method of assessing the
Some of you may not be too keen on mathematics, so here is primary trend. For our example, I selected a 200-day Will-
a method of plotting these bands using common charting soft- iams %R and referred to the standard levels to identify the
ware. Note how some setups result in bands that you are familiar main trend: bullish above -20 and bearish below -80 (another
with, such as Bollinger bands (BB) and Keltner channels (KC). form of hysteresis, if you think about it, in this case applied
These two indicators have long been used to evaluate trend to an oscillator). Readers more familiar with stochastics will
conditions, but if we reduce their usual bandwidth using frac- find that slow stochastics set at 200 days will provide almost
tional deviations, as previously mentioned, they provide excel- identical results.
lent vehicles for applying hysteresis to MA cross-
overs. $DJAAG (Dow Jones-AIG Agriculture Sub-Index) INDX ©StockCharts.com
BBs use a simple MA as the center line and 31-Dec-2007 4:00pm O 80.51 H 81.20 L 80.31 Last 81.00 Chg +0.01 (+0.01%) ▲
SD for the bandwidth, while KCs are con- $DJAAG (Daily) 81.00
MA(20) 79.25 77.5
structed around an exponential moving aver- BB(50,0.8) 73.09 - 75.73 - 78.37
age (EMA) with the width calculated from the 75.0
ATR. In the example presented in this article, 72.5
the slower MA is a simple one (based on an 70.0
arithmetic mean), so BB would be the right << Bearish Bullish >>
67.5
choice, set with the same period (50 bars) and
65.0
0.8 SDs (Figure 4).
By comparing the charts in Figures 1 and 4, 62.5
it is clear that the band has guarded against 60.0
whipsaws and main signals were retained with
57.5
minor delays. Yet these improvements look
modest, and we cannot help wishing for more. 55.0
Can we improve the hysteresis effect?
O N D 06 F M A M J J A S O N D 07 F M A M J J A S O N D
CHOOSING LANES Wm%R(200) -6.80
In the process of applying hysteresis to cross- -20
overs, the role of signal generation originally -50
played by the slower moving average was taken
-80
up by the boundaries of the band. The MA was
left in a central position with the function of O N D 06 F M A M J J A S O N D 07 F M A M J J A S O N D
guiding the band with respect to price evolu-
FIGURE 5: BEARISH/BULL LANES. Using different lanes for different primary trends improves signal
tion. We may have to give back some of the generation. Note how some signals were shifted to the left (earlier) by adopting this differentiated
signaling role to the MA. approach according to prevalent trend conditions.
Signals generation can be

improved by applying
hysteresis in the form of a
Figure 5 illustrates the modified technique. Note how moving band.
some signals were shifted to the left (earlier) by adopting this
differentiated approach according to prevalent trend condi- for practical purposes, you should use a shorter period (less
tions. than 200 bars, or about six months for a daily chart).
MULTIPLE SETS MOVING BANDS

So far we have focused on a single set of two moving averages, Moving average crossovers are lagging indicators prone to
but there is nothing preventing us from using a larger set with false signals. However, the generation of signals can be
multiple bands. We can use, for instance, a pair for short- or improved by applying hysteresis in the form of a moving
medium-term signals and another pair to assess the primary band. By selecting an appropriate lane within the hysteresis
trend. This way we don’t have to use another indicator to band, with respect to the primary trend, you can increase the
evaluate the main trend. The chart of the Standard & Poor’s 500 reliability even more.
in Figure 6 illustrates such an example. You can even apply hysteresis to other indicators and
I used a 12-day EMA as the signaler and two hysteresis bands, construct bandwidths around variables other than volatility,
BB(50, 0.26) and BB(200, 0.13). The pair EMA12/ BB200 was but I will leave that for another article.
used to assess the medium-term trend, and the EMA12/ BB50
duo provided the buy/sell signals. Portugal-based Marco Alves discovered technical analysis
For a longer-term view, the pair BB50(centerline)/ BB200 in 2003 and started to adapt some of the classical indicators
could have been used as a primary trend evaluator as it clearly and design new ones to fit his own investment and trading
relates to the 50- and 200-day MA crossover that is so goals.
commonly used.
The choice for a three-year daily chart resulted from the RELATED READING
need to illustrate the assessment of different primary trends Wilder, J. Welles [1978]. New Concepts In Technical Trad-
and how signals were generated under such different condi- ing Systems, Trend Research.
tions. This kind of chart, however, is not easily readable, and S&C
$SPX (S&P 500 Large Cap Index) INDX ©StockCharts.com

13-May-2008 Open 1404.40 High 1406.30 Low 1396.26 Close 1403.04 Volume 3.3B Chg -0.54 (-0.04%) ▼
$SPX (Daily) 1550
1525
1500
1475
1450
1425
1400
1375
1350
1325
1300
1275
1250
1225
1200
1175
U D U D U D
1150
J J A S O N D 06 F M A M J J A S O N D 07 F M A M J J A S O N D 08 F M A M
$SPX (Daily)
1500
1400
1300
1200
J J A S O N D 06 F M A M J J A S O N D 07 F M A M J J A S O N D 08 F M A M
FIGURE 6: MULTIPLE MA CROSSOVER. The pair EMA12/BB200 (green/red) signaled the medium-term
trend: Up (U) or down (D). The EMA12/BB50 duo established the buy (green arrows) and sell (red arrows)
signals. The price line was cleared from the upper chart. It was plotted on the lower chart for legibility
purposes.

Stocks & Commodities V. 27:01 (10-11, 97): Letters To S&C by Technical Analysis, Inc.
The editors of S&C invite readers to submit their opinions and information on subjects
relating to technical analysis and this magazine. This column is our means of communication FSHOX, first trade
with our readers. Is there something you would like to know more (or less) about? Tell us Pendergast MDD = None
about it. Without a source of new ideas and subjects coming from our readers, this magazine Brown MDD = -18.5% (from 4/19/2006
would not exist. to 6/13/2006)
Address your correspondence to: Editor, STOCKS & COMMODITIES, 4757 California Ave.
SW, Seattle, WA 98116-4499, or email to [email protected]. All letters become the FSUTX, first trade
property of Technical Analysis, Inc. Letter-writers must include their full name and address Pendergast MDD = -0.45%
for verification. Letters may be edited for length or clarity. The opinions expressed in this Brown MDD = -9.5% (from 10/3/2005
column do not necessarily represent those of the magazine.—Editor to 10/20/2005)
CORONA CHARTS AND METASTOCK having problems with it.—Editor My MDD may be off slightly as my
Editor, MOA is slightly different from his (daily
John Ehlers pre- THE KEY TO TREND-FOLLOWING instead of weekly), but I believe his
sented a most in- Editor, MDDs to be incorrect as one does have
teresting discus- Donald W. Pendergast’s article (“The to suffer through daily MDDs. Further,
sion of corona Keys To Trend-Following,” November I don’t understand how an MDD can
charts (“Corona Charts,” November 2008) was interesting and informative ever be “none” for any equity mutual
2008, STOCKS & COMMODITIES) and I but I was surprised by the extremely low fund. There is inevitably a down price
note that while the code for TradeStation maximum drawdowns (MDDs) reported over even short trading periods.
and a few other programs appear in your in his Figure 1, backtested results. A small complaint is that Pendergast
Traders’ Tips section of that issue and I reported on the use of trend-follow- didn’t give the exact trading dates so it
also on your website, a version suitable ing on Select mutual funds in my STOCKS was difficult to replicate his results,
for MetaStock EOD is notably missing. & COMMODITIES January 2005 article, and that there was no April 6, 2007,
Is this because the software is unsuit- “A One Rank Screening Technique For trading day (I used April 5, 2007).
able for the task, or just not available? Mutual Funds.” A key conclusion from Hopefully I am not overlooking some-
TREVOR T. BESTOW that study was that such a trading method thing in my analysis.
Roleystone, Western Australia (the author used a 10-week/40-week NORMAN J. BROWN
MOA) can reduce MDDs by a factor of 2
The Traders’ Tips section contains code (roughly). As the buy & hold for the 14 Donald Pendergast replies:
submitted by the software developers. If Selects the author studied, a significant First off, thanks for reading the ar-
the MetaStock code is missing, it means number have an MDD range from -29% ticle; I am pleased that you found it to be
that the code was not submitted by Equis. (FIDSX) to -60.9% (FWRCX). I would interesting and informative. The draw-
I recommend you visit the MetaStock have anticipated a trading MDD ranging down figure is the maximum drawdown
users forum and ask other users if they from, roughly, MDD -15% to -30%. In from the initial entry price. Here’s an
have been able to recreate the code in fact, his results range from “none” to - example: FNARX trade 2: Date in: May
MetaStock.—Editor 0.13% to the largest at -17.75% (FSAGX 19, 2003 at 9.905. The lowest price
trade 2). subsequent to entry was 9.8623 on July
CORONA CHARTS AND TRADESTATION As I believe those values to be far too 16, 2003, a drawdown of approximately
Editor, low (it would be very nice if they were 0.433%.
Regarding the TradeStation code given true!), I studied four of his funds over Also, regarding MDDs; am I correct
with John Ehlers’s article “Corona essentially the same conditions. I did, to assume you mean “peak to valley
Charts”: When I copied the Easy- however, use a similar MOA of 50/200 maximum drawdown”? I deal with an
Language code for the four corona charts days, as I do not have weekly software. aspect of that statistic in the article as
into PowerEditor in TradeStation 2000i, Here is my maximum MDD found dur- well, comparing the maximum amount
they each failed to verify, indicating ing the buy period compared to his data: of profit attained by the fund with the
that the problem was RGB. Is RGB a actual closed out gain/loss%. Many funds,
function that needs to be installed in FNARX, second trade like the wireless sector fund, gave back
order to use the indicators in Trade- Pendergast MDD = -0.43% a great deal of their open profits, which
Station 2000i? Brown MDD = -20.2% (from 5/10/2006 is why I used that particular ratio
GROVER ERICKSEN to 10/3/2006) (AverageMaxProfit/ AverageTrade) to
alert traders that taking early or partial
“RGB” is a color value you can specify FBIOX, fourth trade profits may be beneficial, depending on
in EasyLanguage. Please check with Pendergast MDD = -2.96% their personal temperaments.
TradeStation directly if you are still Brown MDD = -17.8% (from 2/27/2006
to 5/23/2006)
Stocks & Commodities V. 27:01 (10-11, 97): Letters To S&C by Technical Analysis, Inc.
ANOTHER KEY of a depression-like scenario develop- will rise in a wave I, then fall back in a
Editor, ing. I believe that to be correct as well. wave II by October 2010–11. I do not
I found Donald Pendergast’s article in In the last couple weeks, no doubt believe there will be a depression. The
the November 2008 issue of S&C very some time after your article was writ- S&P 500 wave count does not see that.
interesting. Could I find out the formula ten, we tested the 840 level twice on the
of the EMA 10/40 trading system and the S&P 500, and it has held and bounced MORE WAVES
software used? from there. Editor,
UGO SABADINI I am curious to read your present I read the article “My Kondratieff Wave”
Italy ideas on whether you see a probability in the November 2008 issue of STOCKS &
that the C wave is complete and if so, COMMODITIES, and I have to ask: How
Donald Pendergast replies: would you consider that a new cycle can you say we’re through with winter,
The formula is in the article: Buy on the will begin. What is your expectation and that winter and spring were only three
first daily close after a weekly 10-40- after the C wave completes? years and we’re already into summer? It
seems like we’re still squarely in winter
EMA cross to the upside. Sell on the first C. DOUGLAS WALTERS
with deflation and debt still needing to be
daily close after a weekly 10-40 EMA
Koos van der Merwe replies: washed out of the system. It’s the shortest
cross to the downside. Here’s the
A wave C can fall in a five-wave spring and winter on record. How do you
MetaStock code for the 10-40 E MA make that determination?
crosses only: impulse pattern, or a three-wave ABC
pattern, the latter usually if the wave B JAMES DAVIES
ENTRY Signal: is above the wave 5, which is the case

Koos van der Merwe replies:
Cross( Mov( C,10,E), Mov( C,40,E)) with the monthly S&P 500 chart. Ac-
My Elliott wave count shows that the
cording the S&P 500 chart, therefore,
major C wave is bottoming over the com-
EXIT Signal: the C wave could well be a flat — that ing weeks, probably as President Barack
Cross(Mov(C,40,E),Mov(C,10,E)) is, it won’t fall much lower. Obama settles in. This means that the
However, the count for the Dow Jones bear market will be over, because the C
KONDRATIEFF WAVES Industrial Average (DJIA) suggests a wave is followed by a wave 1 up. How-
Editor, different picture. The C wave has only ever, this does not mean that the recession
I read the recent article by Koos van der completed a wave A down, suggesting is over, because a wave 2 can be a 72%
Merwe, “My Kondratieff Wave” (No- a wave B up before a final wave C retracement of wave 1. The wave 2 bottom
vember 2008, STOCKS & COMMODI- down. Either way, I am expecting a could be, and I believe it will be, only in
TIES), and found it most interesting. I move up between now and April–May, October 2009, which is when the reces-
noticed the Elliott wave chart of the then a further down into October 2009, sion low will occur. It will take another six
Standard & Poor’s 500 projects the C hopefully the bottom of the C wave. I months after that low before the “ex-
wave down to about 800. The author only see the end to the recession in perts” on CNBC announce that the reces-
also indicated that a failure of this level, 2011–12, that end being the low of a sion is over. I am not an economist. I am
which roughly corresponds to the 2003 wave 2 in a new bull market. In other a chartist, and I read my charts. This is
lows, would suggest a high probability words, from October 2009, the S&P what they are telling me.
S&C

Stocks & Commodities V. 27:01 (8): Opening Position by Jayanthi Gopalakrishnan
January 2009 • Volume 27, Number 1

OPENING POSITION
The Traders’ Magazine TM
EDITORIAL
[email protected]
W
Editor in Chief Jack K. Hutson
Editor Jayanthi Gopalakrishnan hen a market is as bearish as it was in
Managing Editor Elizabeth M.S. Flynn
2008, you are tempted to parade
Production Manager Karen E. Wasserman
Art Director Christine Morrison
technical analysis in front of all the funda-
Graphic Designer Karen Moore, Wayne Shaw mental analysts who swear by the buy & hold
Staff Writers Dennis D. Peterson, Bruce Faber strategy, or encourage people to start accu-
Webmaster Han J. Kim
mulating stocks or mutual funds now while
Contributing Editors John Ehlers,
Anthony W. Warren, Ph.D. they are relatively cheap. But does “relatively
Contributing Writers Don Bright, Thomas Bulkowski, cheap” valuation mean the bear market is
Martin Pring, Adrienne Toghraie
over? Every time I look at my charts after I
hear someone mention “capitulation” or “time
OFFICE OF THE PUBLISHER to buy,” all I see is a market that is trending down. And the idea of buying is far,
Publisher Jack K. Hutson far away from my mind.
Credit Manager Linda Eades Gardner
Industrial Engineer Jason K. Hutson
While it is true that you may find some good values or bargains, as long as that
Project Engineer Sean M. Moore trend is pointing down, it is likely that what you are doing is trying to catch a falling
Accounting Assistants Agnes Dimaano knife. What I see now (mind you, I’m writing this toward the end of November
Controller Mary K. Hutson
2008) are markets that are very close to a significant support level, which, if it
ADVERTISING SALES breaks to the downside, could be very ugly. It could be so ugly that I am not even
4757 California Ave. S.W.
Seattle, WA 98116-4499 going to say how far the broad markets could go down. Right now, I’m just hanging
1 206 938-0570 Fax 1 206 938-1307
[email protected] on to the edge of my seat watching the markets to see if that support level breaks
National Sales Manager Edward W. Schramm or holds. At the same time, I’m listening to the talking heads on TV talking about
Classified & Web Sales Chris J. Chrisman a turnaround or a capitulation, or that there is always a bull market somewhere.
Production Coordinator Karen Moore
CIRCULATION
Subscription & Order Service 1 800 832-4642
1 206 938-0570 Fax 1 206 938-1307
[email protected]
Subscription Manager Sean M. Moore
Subscription Sales Karen Adams-Thomas,
S etting aside technical analysis, which most from the fundamental camp
consider voodoo, the economic fundamentals are far from rosy. We have had
a housing collapse, a crisis in the financial markets, a recession, a collapse in the
Teresa Shockley, Carmen Hale auto industry, high unemployment … and who knows what is still to come. All
WEBSITE leading economic indicators are pointing down, just like the markets are.
http://www.traders.com In fact, according to Marketwatch.com, in just over one year (October 2007–
Staff members may be emailed through the Internet
using first initial plus last name plus @traders.com November 2008), the Standard & Poor’s would have lost more than $6.69 trillion
in market cap. Think of how long that will take to recover. What is unfortunate is
Authorization to photocopy items for internal or personal I think that amount will increase in the near term. I think that’s because there are
use, or the internal or personal use of specific clients, is
granted by Technical Analysis, Inc. for users registered with a lot of hedge funds out there that need to liquidate their assets, and I can foresee
the Copyright Clearance Center (CCC) Transactional Report-
ing Service, provided that the base fee of $1.00 per copy, plus
that selloff being more than $500 billion.
50¢ per page is paid directly to CCC, 222 Rosewood Drive,
B
Danvers, MA 01923. Online: http://www.copyright.com. For
those organizations that have been granted a photocopy
license by CCC, a separate system of payment has been y the time you read this, most of these issues would have already taken place
arranged. The fee code for users of the Transactional
Reporting Service is: 0738-3355/2009 $1.00 + .50.
and we may have a more concrete idea of what to expect in the near future.
Subscriptions: USA: one year (13 issues) $64.95; But one thing I’m certain of is that the economy still will be in bad shape. It will take
Foreign surface mail add $15 per year. Air mail: Europe add
$25.50 per year; elsewhere add $39 per year. Single copies a long time to heal this economy. But the bear market may be over well before the
of most past issues of the current year are available prepaid
at $8 per copy. Prior years are available in book format
economy recovers. The charts will tell you when the bear market is over. Keep
(without ads) or from www.traders.com. USA funds only. watching them, because they don’t lie.
Washington state residents add 8.9% sales tax. VISA,
MasterCard, AmEx, and Discover accepted.
Subscription orders: 1 800 832-4642 or 1 206 938-0570.
Technical Analysis of STOCKS & COMMODITIES™, The
Traders’ Magazine™, is prepared from information be-
lieved to be reliable but not guaranteed by us without
further verification, and does not purport to be complete.
Opinions expressed are subject to revision without notifi-
cation. We are not offering to buy or sell securities or
commodities discussed. Technical Analysis Inc., one or
Jayanthi Gopalakrishnan,
more of its officers, and authors may have a position in the Editor
securities discussed herein.
The names of products and services presented in this
magazine are used only in an editorial fashion, and to the
benefit of the trademark owner, with no intention of infringing
on trademark rights.

Stocks & Commodities V. 27:01 (42): Q&A by Don Bright
Q&A
SINCE YOU ASKED
Confused about some aspect of trading? Professional trader Don Bright of Bright
Trading (www.stocktrading.com), an equity trading corporation, answers a few of
your questions. To submit a question, post your question to our website at http://
Message-Boards.Traders.com. Answers will be posted there, and selected questions
will appear in a future issue of S&C.
OTHER OPENING STRATEGIES? ON THE RIGHT PATH? Don Bright of Bright Trading
Are there similar strategies to the open- Sorry if all this has been covered, but I
ing-only strategy for the close of the mar- need to ask again. I have approximately imbalances, then last Friday I wouldn’t
ket each day? Anyone care to share the 30 stocks on my list, all high-volume have sent my MOC orders for PSA and
idea behind them or give me a place to names. I am enveloping based on fair MET, which killed me on the prints for
research if you don’t care to share? value. I am using no-stop, basically stop- $1,000 and $500. Those were my losers
Thanks. —taclander ping manually, crutching or trading out for the day. Can I ask what trading plat-
First off, let’s discuss market mechan- of losers, but I have a 25-cent stop in my form you use to get the 20-second up-
ics at the end of the trading day. All-day- head (adjustments for some stocks, plus dates? Maybe I can come to the training
long orders are submitted to the New or minus 25 cents). For retracements, I camp next time. —ericyyy
York Stock Exchange (NYSE) marked as use six cents for 50% of the position and Having the updates during the last 20
“market on close” (MOC). These orders 25 cents for the second 50% (some ad- minutes of the day (every 20 seconds or
will be given the last trade price of the justments based on stocks). I have, at so) has helped considerably. We see how
day. These orders to buy or to sell are times, used a hard 25-cent stop as well. the flow of the MOC orders go, even to the
matched with each other until the 3:40 Am I at least on the right track? Should point of reversing at 3:50 pm (this is
pm cutoff for this type of order. At 3:40, I expect to make money using this (as I get really helpful). The regulatory entering
the NYSE publishes the MOC imbalances, better and continue to adjust)? What of actual MOC orders is based on the 3:40
which are the excess buy or sell number should an “opens” trader need as a win and 3:50 published numbers, even if they
of shares with this order designation. If rate, win % vs. loss % to be successful? I reverse, but the flow of shares helps a lot.
there were one million shares to buy GE bet many beginners get shaken out with We use the Goldman Sachs RediPlus
market on close and three million shares the “up one day, down two, up three, platform, which has been excellent (espe-
to sell MOC, there would be a negative down three” swings. Is this expected? cially since GS took over a few years ago).
two million excess number of shares to Any help for a newbie is greatly appreci- And feel free to come by, for formal
sell at the last price of the day. The NYSE ated! —VinMan training or just to see how all this works.
will only accept offsetting orders after One major change we’ve made over
3:40 to help cover the excess imbalance. the last few years (due primarily to the WHAT ABOUT PREMARKET TRADING?
The NYSE will republish the adjusted higher volatility) is that we will place I have been doing a lot of premarket
imbalance number at 3:50. This may bring only buys (or sells) when the market is trading lately and have had mixed re-
the number down to one million or so to opening down considerably (or up con- sults. What do you guys think about
sell, or even reverse to an excess to buy. siderably for sell short orders). This way, premarket trading? I trade the stocks that
This information is extremely valu- we don’t end up buying a weak stock, have the most volume that morning, so
able to our traders. We can go with the slightly up from the previous closing price they are earnings or big-news plays. Do
imbalance immediately at 3:40 to make (when the general market is opening way you guys trade the stocks with big news/
money on the immediate move that often up), expecting a pullback. If we’re open- earnings or are they too wild and not
takes place, and cover within a few min- ing up some (say, five to eight S&P worth the risk?—Gimp570
utes (or place the offsetting order as MOC). points), then we use a narrow sell enve- It’s pretty tough to make money
We often make several trades on stocks lope and a much wider buy envelope to premarket. Often, you’ll see a stock trad-
that have unusually high imbalance ensure that our bid is well below previous ing up or down, just to open unchanged
amounts and activity. We have devel- closing price. on NYSE or NASDAQ. And those who
oped several black and gray boxes to help The retracement numbers are similar might actually know something about a
automate order entry as well. to what I use. New traders tend to run 70/ particular stock, versus the the rest of us,
This strategy has gotten easier and 30 win-loss ratio, and it goes up from tend to have an advantage. Lack of liquid-
safer since July 2008, when they started to there. We like the “up one day, down the ity can add to the problem as well. An-
update the imbalances every 20 seconds next” for gap openings. This strategy is other concern are the automated programs
or so (on our trading platform). You may everyday, day in and day out; consistency that some employ that will display the
want to take a look at the MOC from the is what we’re looking for. same pennies, but a different whole-dol-
published imbalances (available on vari- lar amount, to trap unsuspecting traders
ous news sources). WHAT ABOUT THOSE MOCS? (this happens more than you might think).
If I had a 20-second update on the MOC S&C

Stocks & Commodities V. 27:01 (19-22): The MEGAN Ratio by Oscar Cagigas
INDICATORS
Which System = Higher Return?
The MEGAN Ratio

Which system will generate more return? Here’s a whether a specific filter, which eliminates trades and
metric that will help you answer that question. therefore profits, is beneficial to the performance of
the system.
THE GEOMETRIC MEAN
M any metrics can evaluate the performance

of trading systems. These include net
In terms of absolute return, the best system is the one
with the best geometric mean. This is because we
want to reinvest the profits in the next trade. The
profit, recovery factor, Sharpe ratio, K-ratio, and arithmetic average does not take into account the
risk-adjusted return. Even though you can use these reinvestment of profits. For example, imagine the
metrics to select the best trading system for you, the following set of seven trades with a starting capital of
criteria for selecting a system tend to be arbitrary. $10,000:
Metrics such as the Sharpe ratio or K-ratio give an PnL Cap
indication about the slope and smoothness of the
$10,000
equity curve but don’t indicate which system is the
$1,000 11,000
best in terms of monetary performance.
-200 10,800
Not only that, keep in mind that the system with the
-300 10,500
best net profit after simulation is not necessarily the
+600 11,100
one that will generate a higher return when you actu-
+800 11,900
ally trade it. Another system with fewer trades in the
-200 11,700
simulation but better statistics could produce more
+400 12,100
trades and superior performance. Real trading results
will differ from simulated trades because different
The final equity is $12,100 after a $2,100 gain. We
systems have different holding periods for trades.
can express each profit & loss as a percentage of the
Those systems that generate more trades in the same
capital available before the trade. We see this in the
period should do better (if the rest stays the same).
third column:
The MEGAN ratio, which stands for the maximum
exponential growth annualized of the equity curve PnL Equity HPRs
of a trading system, is a metric specifically designed $10,000
to highlight the system that generates more returns $1,000 $11,000 1.1
per year when the profits are reinvested, regardless -200 10,800 0.981818182
of the number of trades, holding period, drawdown, -300 10,500 0.972222222
and so on. This is especially useful to determine 600 11,100 1.057142857
800 11,900 1.072072072
-200 11,700 0.983193277
400 12,100 1.034188034
TWR 1.21
by Oscar G. Cagigas
TOM COCOTOS Stocks & Commodities V. 27:01 (19-22): The MEGAN Ratio by Oscar Cagigas

A holding period return (HPR) of 1.1 is a 10% gain and an expression is the exponent in equation (2):
HPR of 0.9818 is a 1 - 0.9818 = 1.82% loss. If you multiply the
HPR, you get the terminal wealth relative (TWR), a multiplica- T*log(geom) (3)
tive factor (see Ralph Vince’s work). In this case, it is 1.21.
This factor is the quotient between final and initial capital. This If you assume that it is always possible to find a stock that
is the number we multiply the initial account by at the end of generates a signal in your system, the maximum number of
the sequence of trades. trades per year is:
The geometric mean is obtained by taking the nth root of the
TWR. N accounts for the number of trades. In this case, we are Max_trades = 252/AvgBarsHeld (4)
going to do the seventh root of 1.21 –> 1.21^(1/7) = 1.0276, or
a geometric mean of 2.76%. where AvgBarsHeld is the average holding period of a trade in
This percentage is important, since it is the number to the system that we are observing and 252 is the number of
multiply each trade by on average. If there are seven trades trading days in a year.
with a geometric mean of 2.76%, you can expect the final Using (4) in (3), we can finally obtain:
equity to be multiplied by 1.0276^7 = 1.21, or a 21% gain.
Note that the average mean is not the number we are looking MEGAN = (252/AvgBarsHeld) * log(geom) (5)
for. In the previous sequence of trades, the average mean is
2.87%. After seven trades with such an average, you would Or
expect a multiplicative factor of 1.2187, which exceeds the real
gain. This is a common mistake. The average mean does not MEGAN = N*log(geom) (6)
take into account the reinvestment of profits.
where N is the maximum number of trades possible to obtain in
THE MEGAN RATIO a year for that system. Of course, this number is determined by
Say you have two different systems to trade. System 1 has a the holding period of your trades.
geometric mean of 3% and holds positions for an average of The trading system with the higher MEGAN rate is the one
three days. System 2 has a geometric mean of 4% and holds that generates more monetary profit per year. The rate repre-
positions for four days. Which is better for real trading? The sents the maximum achievable exponent of the equity curve in
MEGAN ratio will give you the answer. one year when this equity curve grows geometrically.
In theory, system 2 is better since it has the best geometric A system with a MEGAN of 1 multiplies initial equity by
mean. It should generate more profit after the same number of 2.71828 at the end of the year. A system with MEGAN = zero
trades than system 1. But in the same length of time, they do not returns the same initial capital at the end of one year. A system
generate the same number of trades, and you need to take into with a negative MEGAN ratio generates losses.
account the number of trades when making the comparison. The system with the higher MEGAN ratio generates more
The number of trades is already in the calculation of the money when traded, actively taking all the signals. It is better
geometric mean. Therefore, the mean should be enough to in terms of profit. Let’s calculate the MEGAN ratio for each
choose a system. But when you are actually trading, you can system:
have more trades than those in the simulation. In the simula- System 1 System 2
tion, a trade is generated only if a particular set of conditions
applies. The instrument that we simulate avoids some trades in Geom. 3 4
between, but when you’re actually trading, we can scan the AvgHoldDays 3 4
entire market and find a stock, a futures contract, or several MEGAN 2.48 2.47
instruments with a buy signal today.
After all, if we have a profitable trading system with a very System 1 holds the trades for only three days. That means
good geometric mean, the more we trade, the more profit we that the maximum achievable number of trades per year is 252/
obtain. The geometric mean is related to the TWR according to 3 = 84. If you scan the entire market every day that you are not
this expression: invested, you will be able to get no more than 84 trades. This way,
you will be in the market 100% of the time. The system has a
TWR = geom ^ T (1) geometric mean of 3%, which means that on average, you are
multiplying each new trade by 1.003. After 84 trades, you have:
Where T is the number of trades. We can express (1) as
TWR = exp {T * log(geom)} (2) The system with the higher MEGAN
Log is the natural logarithm and exp is the exponential
ratio generates more money when
function of base = 2.71828. The growth rate of the previous traded, actively taking all the signals.

MEGAN = 84*log(1.03) = 2.48 ■ There are 252 trading days in a year.
System 1 has an equity growth rate of 2.48, which means that Advantages of the MEGAN ratio
at the end of the year, you multiply the initial capital by ■ The ratio allows the comparison of different trading
exp(2.48) = 11.94. The capital will grow exponentially with an systems with different number of trades, net profit, and
exponent of 2.48. System 2 has a greater geometric mean, but trade holding periods.
it holds the positions for more time. It is not as efficient as ■ The ratio helps decide whether a filter (that eliminates
system 1. trades) is beneficial to the system.
System 1 is the better of the two systems. It has a greater ■ The ratio takes into account the reinvestment of profits
MEGAN ratio. It produces more profit when traded full time and in its calculation.
compounding the returns. This 0.01 of difference represents ■ The exponential of the MEGAN ratio indicates the maxi-
exp(0.01) = 1.01, a 1% increase in net profit per year compared mum multiplicative factor that we can expect after a
to system 2. year of trading the system.
■ Since it is an exponent, the MEGAN ratio is not affected
CALCULATING THE MEGAN RATIO by the results of a long testing period or large accumu-
The AmiBroker code to calculate the MEGAN ratio is provided lative profits in the simulation.
in the sidebar “Calculating the MEGAN ratio.” AmiBroker ■ The ratio is not a linear metric. It is geometric like the
allows adding a custom metric to the standard statistics. In this growth of equity of a well-designed system. If a system
case we will add the geometric average and the MEGAN ratio. has a MEGAN ratio 0.69 (log(2)) better than another
We can add the code in the sidebar to our system or save it system, then it duplicates its net profit.
in the included folder and just add the following line to the ■ The ratio is useful for systems that generate profits in
system: many instruments (stocks, futures, bonds, and so forth).
#include <geom.afl> Disadvantages of the MEGAN ratio

■ It does not address consecutive losses, drawdown, and
That will add two metrics to the standard statistics, the geomet- so forth (although it is not intended). It only gives an
ric mean, and the MEGAN ratio. indication of maximum profit achievable. For other
measures of comfort, we already have the rest of the
Assumptions metrics.
The MEGAN ratio is based in the following assumptions: ■ In systems specifically designed for a unique instru-
ment (for example, a system for the eurodollar), the
■ You can get more trading signals than in a simulated
MEGAN ratio will be useful for evaluating performance,
environment by searching a database containing more
but its absolute value will not translate into an achiev-
instruments than a simulation portfolio.
able profit since we will not be able to increase the
■ The extra trades (obtained to reach the maximum per
number of trades due to the nature of the system.
year) have the same characteristics (geometric mean) as
the rest.
Given its advantages and disadvantages, the MEGAN ratio is
■ We are working with daily data.
a solution for comparing systems with different characteristics
or testing changes in the same system. The better the absolute
CALCULATING THE MEGAN RATIO VIA AMIBROKER
value of the ratio, the more profit the system will achieve after
MEGAN RATIO CALCULATION a year of trading it.
// geom.afl
// COMPARING TWO TRADING SYSTEMS
SetCustomBacktestProc(“”);
if (Status(“action”) == actionPortfolio)
System 1 is a breakout system of using 20 and 10 periods
{ (Figure 1). The code is shown in the sidebar “Code for system
bo = GetBacktesterObject(); 1.”
bo.backtest(); There is no stop-loss. The only exit is the low of a 10-day
st = bo.getperformancestats(0); period. I applied this system to a portfolio of 30 stocks from
geom = (st.getvalue(“EndingCapital”)/st.getvalue(“InitialCapital”))^(1/
January 1, 2000, to January 1, 2008, in a cash account (no
st.getvalue(“AllQty”));
geomp = 100*(geom-1); //geom percentage
margin) with initial capital of $100,000. No commissions are
mtpy = 252/st.getvalue(“AllAvgBarsHeld”); //maximum number of trades per deducted. The portfolio components are ARBA, AUXL, AUY,
year AVP, BIDU, CBG, CLR, CRM, ELOS, EXM, FMD, FSLR, GM,
meg= mtpy*ln(geom); //max exponential growth rate (annual) GME, ICE, ILMN, INFY, ISRG, LDK, LVS, MR, NITE, NSC, NTY,
bo.addcustommetric(“Geom%”, geomp); NUE, PCP, RIMM, SHLD, UTHR, and WNR.
bo.addcustommetric(“MEGAN Ratio”, meg);
The results after the portfolio simulation are displayed in
}
Figure 2. Net profit is $3.46 million after 77 trades. The
CODE FOR SYSTEM 1 USING AMIBROKER
RIMM - Daily 11/16/2004 Open 27.4325, Hi 27.6523, Lo 27.0196, Close 27.2494 (-0.9%) GS 32.0 //entry//
30.0 Buy = Cond = H > Ref(HHV(H,20),-1);

BuyPrice = Max(Ref(H,-1),Open);
28.0
27.2494
26.0 //exit//
24.0 Sell= L < Ref(LLV(L,10),-1);
SellPrice = Min(Ref(LLV(L,10),-1),Open);
.22.0
20.0 but system 2 trades a lot more. The geo-

metric mean of system 1 is clearly better,
SYSTEM 1 18.0 7.90% vs. 0.94% for system 2. The Sharpe
August September October November ratio of system 2 is better. With the second
system we have a smaller geometric gain
FIGURE 1: THE FIRST SYSTEM. System 1 buys a 20-day breakout of the highs. It sells a 10-day breakout of per trade, but we can trade more frequently
the lows. than with the first. Therefore, each time
we make a trade, we improve our result,
since we multiply the equity by the geo-
Net Exp% MSD% RF CAR/MDD PF Payoff
metric mean. Which one will generate
3,467,665.53 73.26 -43.27 2.24 1.3 2.33 1.95
more profit when traded?
Sharpe # Trades %W Avg% MedGeom% W avg Bar L avg B
The average holding period of all trades
1.04 77 54.55 6.14 4.75 33.3 12.6
(winners and losers) in system 1 is 23.90
FIGURE 2: PORTFOLIO SIMULATION RESULTS
days. The average holding period in system
2 is 5.20 days. This can be calculated from
the statistics; nevertheless, AmiBroker pro-
RIMM - SIST MEGAN2 = 27.25, MA1(SIST MEGAN2,5) = 27.49, MA2(SIST MEGAN2,20) = 2 vides this value directly. The maximum
32.0
number of trades per year (N in equation 6)
is:
30.0
28.4114 Max Trades per Year

28.0
27.4878 System 1 System 2
27.2494 10.54 48.46
26.0
System 2 will give us four times more

24.0
trades than system 1. The MEGAN ratio for
.22.0
the two systems are
20.0
MEGAN Ratio
System 1 System 2
18.0
0.49 0.45
FIGURE 3: SYSTEM 2 SIGNALS APPLIED TO RESEARCH IN MOTION

CODE FOR SYSTEM 2 USING AMIBROKER
//SETUP//
geometric mean of this system is 4.75%. setup=MA(C,5) > MA(C,20);
System 2 is a moving average system that takes long trades
after the crossover if the previous high is exceeded. We sell only
if the low of two previous bars is broken. The code is displayed //Long Entry//
in the sidebar “Code for system 2.” The simulation portfolio
Buy= Ref(setup,-1) AND H > Ref(H,-1);
and conditions are the same as for system 1.
In Figure 3 we can see the signals of system 2 when this BuyPrice=Ref(H,-1);
system is applied to Research In Motion (RIMM). The statis-
tics for system number 2 are displayed in Figure 4. The net
profit is $3 million after 368 trades. The geometric mean is //SELL//
0.94% per trade. Sell=L < Ref(LLV(L,2),-1);
Both systems have similar profit and similar drawdown,
SellPrice=Min(Ref(LLV(L,2),-1),Open);

Net Exp% MSD% RF CAR/MDD PF Payoff
3,057,745.24 60.59 -47.69 6.54 1.13 2 2.49
1.29 368 44.57 2.03 0.94 7.6 3.3
FIGURE 4: SYSTEM 2 STATISTICS
We can see that although we can get four times more trades In Figure 5, we can see the exponential growth of the equity
in system 2, the maximum achievable exponent of the equity curve for both systems. A slight difference in the MEGAN ratio
curve is greater in system 1 because it has a greater geometric translates to a bigger difference in the growth of the equity,
mean. That is not always the case, since the system with the which increases with each new trade. The curve for system 2
has a lower geometric mean but more trade fre-
120 quency. The factor 4.60 is the quotient between
System 1 the maximum number of trades in systems 2 and
System 2 1.
100 In Figure 6 we see the equity curve of both
exp(n*In(1.0475)) systems. This is a real picture of the ideal curves
80 shown in Figure 5. The curve for system 1 is
stepped and it is always possible to make addi-
tional trades during the flat periods.
60
EVALUATING THE EFFECT OF A FILTER
In the previous example, we have seen that sys-
40 tem 2 makes a lot of trades. Perhaps we can add a
filter to try to improve the profit of this system.
20 Once we filter some trades, it is clear that the
remaining trades will generate less total profit
exp(n*4.60In(1.0094)) (less than $3 million). We must know whether
0 this reduced profit generated by less trades is an
1 7 13 19 25 31 37 43 49 55 61 67 73 79 85 91 97 improvement. In other words, we want to make
FIGURE 5: IDEAL CURVES changes to the system and evaluate if these changes
allow the system to generate more profit after a
year.
4000000 We can trade only in the most favorable months.
System 1 In this case, we are going to remove trades from
3500000 System 2 April to July, inclusive. The new filtered code is
seen in sidebar “Effect of a filter.” The result of
3000000
the simulation (same conditions apply) is dis-
2500000 played in Figure 7.
The MEGAN ratio for this filtered system is
2000000 0.62. This means that capital grows at a higher
1500000 rate in the filtered system. So the filter is benefi-
cial for the system, since it will generate more
1000000 profits.
Although the average holding period is the
500000
same (we have removed trades without modify-
0 ing the entry and exit logic), the geometric aver-
04/18/2000
07/31/2000
12/08/2000
04/05/2001
08/03/2001
12/04/2001
04/05/2002
07/05/2002
10/08/2002
02/05/2003
06/04/2003
10/21/2003
01/30/2004
05/19/2004
09/24/2004
01/26/2005
04/15/2005
08/11/2005
11/11/2005
02/27/2006
06/21/2006
10/03/2006
01/08/2007
04/27/2007
07/27/2007
12/04/2007
age is now superior (1.33 vs. 0.94). This is be-

-500000 cause we have removed the less efficient trades.
Date
The improvement in the growth ratio is 0.17. The

FIGURE 6: EQUITY CURVE, BOTH SYSTEMS
new system is far superior even to system 1. Other
kinds of tests like taking only trades if average
directional movement index (ADX) is rising and above 20 or
better geometric mean can be a long-term system that will not waiting for the averages to be rising have not improved the
generate better results than a short-term system with more MEGAN ratio of this system.
modest statistics.
The MEGAN ratio is shown here because of the need of a Oscar G. Cagigas has a degree in telecomunications engineering. He
metric that takes into account performance and time together. is a private trader with more than 10 years of experience. He is the
The difference in MEGAN ratio between these two systems founder of www.onda4.com, a financial website in Spanish. He has
translates to a 4% difference in maximum profit after one year: published three books regarding Elliott wave theory, trading systems,
exp(0.49)/exp(0.45). and money management.
EFFECT OF A FILTER
#include <geom.afl>
//SETUP//
SUGGESTED READING
setup=MA(C,5) > MA(C,20); Vince, Ralph [1995]. The New Money Management, John
//filter//
Wiley & Sons.
filtro=Month() < 4 OR Month() > 7;

//Long Entry//
Buy= Ref(setup,-1) AND H > Ref(H,-1) AND filtro;
BuyPrice=Ref(H,-1); Net Exp% MSD% RF CAR/MDD PF Payoff
2,403,295.53 42.18 -32.28 8.52 1.54 2.44 2.85
//SELL//
Sell= L < Ref(LLV(L,2),-1); 1.57 243 46.09 2.4 1.33 7.9 3.2
SellPrice=Min(Ref(LLV(L,2),-1),Open); FIGURE 7: SIMULATION RESULTS
S&C

Stocks & Commodities V. 27:01 (66-69): Three Rules, One Easy Way To Trade ETFs by Larry Conners and David Penn
TRADING SYSTEMS
Rolling The Double 7s
Three Rules, One Easy Way

To Trade ETFs
last but not least, we lim-
ited the strategy to three
simple rules.
THE DOUBLE 7S
Before we tell you those
rules, first understand a
bit about this trading strat-
egy, which we call Double
7s. We tested this strategy
on both the Standard &
Poor’s 500 (SPX) and the
NASDAQ 100 (NDX) from
January 1995 through
April 2008. The results
over that time period for
the S&P 500 were +1,133
points, with nearly 80%
of the trades closing prof-
itably. In the NASDAQ
100, the results were simi-
larly impressive. From
January 1995 through
April 2008, the Double
7s strategy in the
N ASDAQ 100 yielded
2,822 points, with
79.84% of the trades clos-
ing in the green.
BORIS LYUBNER
We also tested the
Double 7s strategy on in-
ternational ETFs such as
the FXI, representing the
Chinese F TSE /Xinhua
Index, and the EWZ, rep-
Here’s a short-term trading strategy to trade exchange traded funds. resenting the Brazilian
stock market. Again, the
by Larry Connors and David Penn simulated trading results
spoke for themselves. In
famous trader once remarked that he could publish the secrets to his winning trading the FXI, we saw 73% cor-
A strategy in the newspapers and no one would follow them. The key, this famous
trader explained, was consistency and discipline.
In our opinion, when it comes to trading and trading strategies, one of the things
that makes consistency easier and discipline a bit less daunting is simplicity. With
rect trades since the ETFs’
inception. In the EWZ, the
percent of correct trades
was more than 80%.
few exceptions, all the trading strategies we have developed over the years have had These were long-only
simplicity as their hallmarks. trades over a period that
What we’ve developed in this latest trading strategy is short-term trading at its simplest. Instead included a boom, a bust,
of stocks, we focus on the smaller universe of market index exchange-traded funds (ETFs). We a low-volatility recov-
also created the system as long-only — no need to worry about borrowing shares of ETFs. And ery, and then another
S&P 500 Index 806.58 -52.54 -6.12%
D: 11/13/08 O: 138.58 H: 1387.61 L: 1378.80 C: 1384.42 Y: 1291.47
Simple Moving Avg. (200) Last=1249.62
1440
1430
C
1420
bust. Note also the high win rate. This is a character- A 1410
istic of our trading systems, one that traders who 1400
have used our strategies have come to appreciate. 1390
Trading systems with high win rates tend to make 1380
the trading experience much more enjoyable for the 1370
trader. This is because the experience of trading is, B 1360
by and large, a positive one since traders are making 1350
money on the majority of trades. 1340
This also speaks to the issues of consistency and 1330
discipline we’ve mentioned. When the day-in, day- 1320
out experience of trading is largely positive, the 1310
discipline of following the rules and making your 1300
trades is easier to abide by. Over time, even the most 1290
profitable trading systems can be hard to adhere to if 1280
PROPHET
the trader feels like every trade boils down to a 50/ 10/23 10/30 11/06 11/13 11/20 11/27 12/04 12/11 12/18
50 crapshoot — or worse.
FIGURE 1: DOUBLE 7’S STRATEGY TRADE — SPX. The S&P 500 made a new seven-day high
at point A. After moving higher for another six days, the S&P 500 topped and began to pull back,
THE THREE RULES creating a new, seven-day low at point B. We take a long position at the close. Five days later,
If you’re still looking for a profitable trading system the S&P 500 made a new, seven-day high at point C. We exit our long position on the close.
for exchange-traded funds that is easy to follow and
has win rates of nearly 80%, then it’s time to intro-
duce you to the three simple rules that make up our Double 7s
strategy:
DOUBLE 7S STRATEGY IN ACTION
1 The ETF is above the 200-day moving average. The SPX made a new, seven-day high on November 14, 2006
2 The ETF closes at a seven-day low. Buy on the close. (Figure 1). Eight trading days later, the SPX pulled back
sharply, setting a new seven-day low.
3 Exit when the ETF closes at a seven-day high.
The Double 7s strategy has us buy the SPX, in the form of
the SPY or other S&P derivative such as the 2X leveraged
And that’s it. We hope we didn’t disappoint you if you
ProShares Ultra S&P 500 exchange-traded fund (SSO) on the
were looking for something more complicated, but we be-
close of the new, seven-day low on November 27.
lieve in letting the market themselves do all the hard work.
We hold that position until we get a rally that lifts the S&P
The Double 7s strategy is a mean-reversion, swing trading
500 to a new, seven-day high.
method of trading ETFs.
That rally began almost immediately and, five days later on
By separating out only those ETFs that are above the 200-day
December 4, the S&P 500 was making a new, seven-day high.
moving average, we focus on those stocks that are statistically
The Double 7s strategy has us sell our position on the close
more likely to move higher after short-term pullbacks. This
of this day. The Sso trade would have been just as impressive,
observation has been at the core of our successful short-term
if not more so when considered on a percentage basis.
trading for more than a decade and is the first and most
Here’s an example in December 2007 with the NASDAQ
important rule in the Double 7s strategy.
100 (Figure 2). Here our market proxy is the QQQQ. Trading
Everyone wants to buy short-term pullbacks in uptrends.
above its 200-day moving average, the Q made a new, seven-
But many traders get caught up trying to devise a perfect
day high on November 28. It moved higher fitfully before
method to deduce when a pullback has sufficiently pulled back.
peaking and selling off sharply intraday on December 11,
In our testing of the Double 7s strategy, we found positive
2007.
results with a number of time frames but discovered that the
Four days later, with sellers continuing to overwhelm
best edges were in ETFs that had made seven-day lows.
buyers, the NASDAQ 100 made a new seven-day low. The
The compliment to buying on a pullback is to sell into
Double 7s strategy alerts us to this buying opportunity and
strength. For years, professional traders have insisted that
has us buy the Q on the close of the new seven-day low on
“when the ducks are quacking, feed them,” preferring to sell
December 14, 2007.
their positions into strong demand instead of waiting for that
After two days of relatively constricted sideways trading,
demand to subside. We have used a variety of different
the Q explodes to the upside and sets a new, seven-day high
techniques and tools — all simple and straightforward — to
within five days of our initiating the trade. That is our exit,
determine when that demand was suitably strong. With the
and as of the close on December 21, the position is sold for a
Double 7s strategy of short-term trading in ETFs, exiting
profit.
trades at a seven-day high tested out to be an ideal compli-
Assuming we used the Q as our NASDAQ 1000 proxy, we
ment to the entry at a seven-day low.
POWERSHARES QQQ TRUST SENIOR 1 ETF-(OTC) 26.86 -1.48 -5.22%

D: 11/09/07 O: 50.6022 H: 50.9913 L: 49.8741 C: 49.8741 R: 1.1172 Y: 55.15
There are plenty of complicated
Simple Moving Avg. (200) Last=43.03
55
ways of making money in the
C
markets. The Double 7s strategy
54
B is not one of them.
A 53
itself readily to mechanical trading for those who
52 prefer to trade using fully automated systems. The
strategy can even be used as a part of a market
51
timing strategy for trading individual stocks and
50 options when the market as a whole is down and
stocks more likely to be on sale.
49 Most important, the Double 7s strategy lets
traders take advantage of market psychology with
48 every trade. We are buying when fear is reaching
extreme levels in the short term. We are selling
47 when greed is the most powerful driving emotion.
This is how professional traders take advantage of
46 what is referred to as the “wrong-way crowd” in
11/12 11/19 11/26 12/03 12/10 12/17 12/24 12/31 01/07
trade after trade.
FIGURE 2: DOUBLE 7S STRATEGY TRADE — QQQQ. The Q made a new seven-day high at point There are plenty of complicated ways of mak-
A. The ETF worked its way higher over the next several days before selling off and pulling back to ing money in the markets. The Double 7s strategy
a new, seven-day low at point B. We take a long position at the close. Again, within five days, the
Q made a new, seven-day high at point C, where we take profits on the close.
is not one of them. By sticking with ETFs that are
above their 200-day moving average, buying them
would have taken our long position on the close of December 14 and on short-term weakness and selling them into
exited that trade on the close of December 21 for a gain of more than 2%. short-term strength, we have created another high
More aggressive traders can take the Double 7s strategy further by win rate, straightforward mean reversion swing
using options on the SPY and Q. Traders can also use the leveraged trading system that we think more and more trad-
exchange-traded funds offered by Rydex, ProShares, and other funds — ers can benefit and profit from.
although, as with options, traders should be aware that additional
leverage means a greater potential for loss as well as profit. Larry Connors is C EO and founder of
One of the more interesting aspects of the Double 7s strategy is that TradingMarkets.com. David Penn is editor-in-
traders can still trade the strategy during the bear market by trading chief of TradingMarkets.com.
inverse ETFs. Exchange-traded funds like the SDS and the QID provide
both 2x leverage and an inverse relationship to the SPY and Q. As such, RELATED READING
when the $SPX and $NDX are untradable because they are below the Connors, Larry, and Conor Sen [2004]. How
200-day moving average, the SDS and QID — which as inverse ETFs Markets Really Work: A Quantitative Guide To
will be above the 200-day moving average — can be used as a Stock Market Behavior, TradingMarkets Pub-
substitute. lishing Group.
Connors, Larry, and Cesar Alvarez [2008]. Short
SIMPLE AT BEST Term Trading Strategies That Work: A Quan-
The Double 7s strategy is easy to follow. Traders can use price channels tified Guide To Trading Stocks And ETFs,
to help mark where the seven-day highs and lows occur, letting breaches TradingMarkets Publishing Group.
of those levels signal entries and exits. The Double 7s strategy also lends S&C

Stocks & Commodities V. 27:01 (84-89): Traders’ Resource: Trading Systems by Technical Analysis, Inc.
Trading Systems
Trading systems can help take the TRADERS'
RESOURCE
subjective interpretation out of trading
decisions by providing automated buy and TOP 10 VIEWED
sell signals based on preprogrammed rules. LINKS
Trading systems are usually computer TRADING SYSTEMS
programs but can also be a real-time service
issuing signals or a set of published rules to Company Trading System
follow. They can rely on one or more
trading disciplines, such as artificial intelligence, Gann analysis, 1. Profitunity Trading Group Profitunity Trading Group
astrology, indicator sets, or custom rules. This month in Traders’ 2. Key-Volume Strategies Inc. Options 40 In 4
Resource, we present a list of publicly available trading systems, 3. AbleSys Corporation AbleTrend 7.0
most of which are sold as software packages for personal 4. Investment Research Co. E-Mini Russell 2000 Trading System
computers. 5. National Trading Group Winning Edge Trading System
We contacted trading system developers and vendors and 6. TrendPro, Inc. Floor Traders Tools 8.0
asked them to fill out an online survey form to describe their 7. Gecko Software, Inc. Bulls ‘n Bears Red Light,
product or products. Here, you’ll find an abbreviated list of the Green Light System
products. While we’re only able to present a small portion of 8. TradersCoach.com Applied Reality Trading (ART)
the information here in the magazine due to the space constraints 9. Forex Capital Markets LLC Forex System Selector
of the printed page, you’ll find the complete listing of product 10. DEL Associates Ltd. CupTrade Strategies
specifics at our website, Traders.com. Just click on the Traders’
Resource link and then on the Trading Systems category link. These are the 10 trading systems clicked on most often on the
Traders’ Resource website. Each company is listed in order of
The product information you’ll find at our website for clicks received. This is not an editorial rating, ranking, or opinion.
trading systems will help answer questions such as: What For more information on specific products and services, try
indicators does the system utilize? What markets are followed checking store.Traders.com for archived S&C product reviews.
by the trading system? Are additional applications required to
use the trading system? What types of customer support are TRADERS’ RESOURCE AT TRADERS.COM
offered? What are the primary features of the trading system? In addition to the trading system listing at Traders.com, you’ll
While we are unable to present track records or rankings in the also find listings of brokerages, mutual funds, data services,
listing, we hope this resource will help you narrow down your courses and seminars, software, and more. We hope this will
search toward finding the right trading system. help you learn about products that may help your trading.
Company Website Trading System Backtest?
2HEDGE Company www.2hedge.com NeoTicker + McSpread Application Yes

3F Forecasts www.3fforecasts.com 3F Forecasts No
AbleSys Corporation www.ablesys.com AbleTrend 7.0 Yes
Advanced Trading Systems Inc. www.trading-systems.info Atlas-35 RT; Chronos-1 Yes
Advantage Systems www.advantage.online.pl Advantage System Yes
AfterHourTrades.com, Inc. www.afterhourtrades.com AfterHourTrades.com Yes
All In The Money allinthemoney.com All In The Money Yes
Allegro Trading Systems www.allegrotrade.com Allegro Trading Systems Yes
Alpha Trading Systems www.alphafin.com Global Futures System No
AmCan Financial Group www.amcan.net AmCan PSI3WM_Sys Yes
Aminda, LLC www.aminda.com EquityTrading System Yes
Analysis & Knowledge www.analysis-knowledge.com Patterns Yes

ATRADE Investment Technologies, LLC www.findmynexttrade.com ATRADE Online - Visually Simplified Stock Alerts Yes
Attain Capital Management www.AttainCapital.com Forex Trading System Execution Yes
Autumn Investments, Inc. www.mikeysmethods.com Trading Commodity Futures Yes
Blue Wave Trading Research and Dev. www.bluewavetrading.com BWT Zones SP Yes
Bon Trade Solutions bon-trade.com Bon Trade Direct Access Yes
BrainTrading Solutions Inc. www.BrainTrading.com BrainTrading System No
Breakout Futures www.BreakoutFutures.com MiniMax Yes
Ca$h Cow Trading Strategies www.cashcow-windicator.com WINdicator Trading Strategy Yes
Calsoft www.californiasw.com DeRisk 3.1 No
Cap Portfolios www.capprotection.com/ CAP Securities Portfolio Yes
Capital Essence Corp. www.CapitalEssence.com Mai Oscillator Model Yes
Channel Trend Inc. www.channeltrend.com Channel Trend Yes
Chaos Enterprises www.ChaosFutures.com ChaosFutures S&P 500 Futures Thrust Forecasts Yes
Chapman Advisory Group www.stockconsultant.com StockConsultant.com Yes
Choice Day Trades www.choicedaytrades.com RPM - Reliable Pattern Match Yes
Chuck Le Beau’s System Traders Club traderclub.com Various trading systems for futures markets Yes
Circular Logic, Inc. www.circularlogic.com Circular Trading System No
Clayburg.com www.Clayburg.com Belly Trader; Feeder Trader No/Yes
Cleveland SOHO Enterprises www.take-overs.com/ Insider’s eye Yes
my_commod.htm
Cobra Trading, Inc. www.cobratrading.com Instaquote No
Commodity Systems, Inc. (CSI) www.csidata.com MULTI MARKET ANALYZER Yes
Compass Financial Futures and Forex www.compassfinancial.com Forex Trade Oracle No
Creative Breakthrough, Inc. www.traderassist.com %C Day Breaker; others Yes
Creative Investment Research Group www.SmartTrading.com Ryan Jones Live Account - Actual fills No
Data Shaping Solutions, LLC www.datashaping.com/ Datashaping Yes
DataCode Inc. www.datacodeinc.com The Netfeed System Yes
Dave Fox, CTA www.dollartrader.com DOLLAR TRADER Yes
Day Traders Win daytraderswin.com DayTradersWin Trading System Yes
Decision Software www.dsoftware.com DSTS / BondOne Yes
DEL Associates Limited www.cupwatch.com CupTrade Strategies Yes
Dipl.-Ing. Volker Butzlaff www.zentrader.de Zen Trading System v1.53 Yes
DOLLAR TRADER www.DollarTrader.com DOLLAR TRADER for the Currrencies Yes
Econom-Expert Ltd. stockfusion.net ForeStock Yes
Edge Ware, Inc. www.edge-ware.com FastBreak & FastBreak Pro Yes
eFloorTrade www.efloortrade.com CyberMint Yes
EFT Capital (Pty) Ltd. www.eftcapital.com RTTC - Real Time Trade Communicator No
Financial-edu.com www.financial-edu.com AnalyzerXL - Excel Trading System; others Yes
Flash Futures www.flashfutures.com SysTrak No
Flow of Capital Inc. www.flowofcapital.com Proactive Timing System No
Floyd Upperman & Associates www.upperman.com Floyd Upperman and Associates Yes
Foresight Investment Strategies, Ltd. www.foresight-investment.com/ Private Forex Managed Account Trading System Yes
forex_trading.php
Forex Capital Markets, LLC www.fxcm.com Forex System Selector; FXCM Trading Platform Yes
Forex Trade Signals www.fxts.com Forex Trade Signals Yes
Formula Research, Inc. www.formularesearch.com Multiple timing models Yes
Frank Petricca Inc. www.soybeanmarket.com LTPS - Long Term Profit Strategies Yes
Futures Trading Systems by Trendchannel www.FuturesTradingSystems.com Capitalflow Porfolio by Trendchannel Yes
FuturesExaminer.com www.FuturesExaminer.com FuturesExaminer.com Yes
FXDirectDealer www.fxdd.com FXDirectdealer Yes
TRADING SYSTEMS. Look for the complete listing at Traders.com under Traders’ Resource.
FX-Strategy.com www.fx-strategy.com FX-Strategy Trading System Yes

Gannline.com www.gannline.com TIME TARGET Yes
GATE Tecnologie Informatiche S.r.l. www.gatelab.com GATEMate Yes
Gecko Software, Inc. www.geckosoftware.com Bulls ‘n Bears Red Light Green Light System Yes
Global Forex Trading (GFT) www.gftforex.com/ Facilitated Systems Yes
GlobalTec Solutions, LLP www.globaltecsolutions.com Wizetrade 4X Made Easy & Options Made Easy Yes
GorillaTrades, Inc. www.GorillaTrades.com GorillaTrades Yes
Hawkeye Traders (Home of www.HawkeyeTraders.com Hawkeye Traders No
Volume Spread Analysis)
Highlight Investments Group QQQ-Options-Trading.com QQQ-Options-Trading Yes
HotSignals.com www.HotSignals.com HotSignals.com Yes
Hotspot FX, Inc. www.hotspotfx.com Hotspot FX Yes
Hottinger Capital Consulting, Ltd. www.HottingerSignals.com Hottinger’s E-Zone Signals No
INO.com www.ino.com MarketClub Yes
InstantCharts www.instantcharts.com Ultimate Trading System Yes
IntelliTrade Labs www.Intelli-Timer.com Intelli-Timer.com Yes
InvestIN Securites Corp. www.directaccesselite.com DirectAccessElite Yes
Investment Research Company www.investmentresearch.biz E-Mini Russell 2000 Trading System; others Yes
Investors Alley Corp. www.investorsalley.com Trader’s Tribune Swing Trading Service Yes
Joe Krutsinger, CTA www.eTrackRecords.com TIME ANY Yes
John Craciun www.MarketTide.com Trading System Testing Software (TSTS) Yes
Julian’s Mutual Funds Tactics No website Randy Julian’s Proprietary Trading System Yes
Just Data www.justdata.com.au ScanVest Yes
Key-Volume Strategies Inc. www.key-volume.com Options 40 In 4 Yes
Kingsbury Financial Management, Inc. www.tradersparadise.com TradeScan Trading Alerts No
Lighthouse Trader LLC www.lighthousetrader.com North Star Trading System No
Lind-Waldock www.lind-waldock.com Lind Auto-Execute Yes
longtermtrading.com www.longtermtrading.com WaveRider; others Yes
Lucky Tips www.luckytips.co.uk The ZRC System; ZTL Forex Trading Metho Yes
Market In&Out www.marketinout.com/ Trend Lines Breakout; others No
MarketDirection www.marketdirection.net Genesis; JJ Yes
Marketline Advisory Service personalpages.tds.net/~ Universal Reversing System; others Yes
stantam2/index.htm
Masters ‘O’ Equity Asset Management www.mastersoequity.com Star Trading System Yes
Mathematicus Labs www.mathematicuslabs.com TradeStation Indicator Library Yes
Matheny Enterprises www.ment.com ASL Library for TradeStation; others Yes
MESA Software www.mesasoftware.com MESA Bonds Yes
Meyers Analytics LLC www.meyersanalytics.com Key Daily & Intraday Trading Systems v5; others Yes
MG Financial Group www.mgforex.com DealStation No
MHP Systems Inc. (Chartfilter.com) www.chartfilter.com ChartFilter Stock Tools No
Micro-Labs, Inc. www.microlabs.com/trade MoneyVolume No
MicroMedia www.cashinonchaos.com Individual XGO Charts No
Mindfire Systems mindfire-systems.com Catscan 4; others Yes
MomentumT www.stressfreetrading.com Momentum Share Trader Yes
MoneyMachine, Inc. www.moneymachinenet.com MoneyMachine Yes
MTPredictor Ltd. www.mtpredictor.com MTPredictor No
Murphyware www.ontheclose.com Del Mar Yes
National Trading Group www.winningedgesystem.com Winning Edge Trading System Yes
New Era Trader www.NewEraTrader.com trading magic Yes
Nexgen Software Services nss-t3.com T-3 Fibs Protrader Yes
NexTrend Technologies, Inc. www.nextrend.com NexTrend Electronic Direct Access Trading (EDAT) No
NobleTrading.com www.nobletrading.com RealTick III Yes
OceanBlue Publishing www.mayyoubehappy.com/oceantheory.html Ocean Master Package Yes
Omicron Research Institute www.omicronrsch.com QuanTek Yes
Omnifund Limited www.findatanet.com Streamlined Market Optimizer III: The Master Key No
OnTick www.OnTick.com QuoteMachine Yes
OPTiMAX de.geocities.com/fh24/opti.html OPTiMAX-FIBONACCI Trading System Yes
OptionScan Ltd. optionscan.googlepages.com OptionScan Yes
OptionVine www.optionvine.com OptionVine No
OptionVue Systems International Inc. www.optionvue.com OptionVue 5 No
PageTrader www.pagetrader.com PageTrader Yes
Petros Development Corp. www.andromedafutures.com Andromeda Trading System; others Yes
PFSoft www.pfsoft.com ProTraderFX Yes
PIMS, Inc. www.pimsva.com Currency Futures Options Trading System - CFOTS Yes
Pivot Research & Trading Co. www.PivotTrader.com Daily Bias Trading System No
Platinum Trading Solutions www.platinumtradingsolutions.com Platinum Commodity Spread Program Yes
Practical Trading www.practicaltrading.com Practical Day Trading System Yes
Prime-Line www.Prime-Line.com Auto Prime-Line; others Yes
PrimeMarketSignal www.PrimeMarketSignal.com PrimeMarketSignal INDEX Yes
PRISM Trading Advisors, Inc. www.prismtrading.com PRISM MACD Trading Plan & Manual Yes
Profitunity Trading Group www.Profitunity.com Profitunity Trading Group No
ProSticks Financial Solutions www.prosticks.com ProSignals Trend Scanner Yes
Pure Trading Systems ForexKillerMoney.com Forex Killer No
Quantitative Equity Research www.quantequityresearch.com Quantitative Equity Research Trading System Yes
RealityTrader.com www.realitytrader.com realitytrader.com Yes
Robbins Trading Company www.robbinstrading.com System Assist Yes
Ryan Jones www.SmartTrading.com S&P Probability Calendar - Ryan Jones No
Scientific Systems, Inc. www.scientific-systems.com TradeMaster (in several versions) Yes
Shield-Systems Technology Co. www.shieldsys.com Phenix Forex automated trading system Yes
Skilltrader, LLC www.stocktrading.nu SkillTrader Yes
SmarteTrades.com www.SmarteTrades.com Buffet Equity System; others Yes
SmartProj www.smartproj.com StraTrader Yes
Spooz, Inc. www.spooz.com Fractalz3 Yes
Steve Nison’s Candlecharts.com, Inc. www.candlecharts.com Nison Real Time Candlle Finder Yes
Straiton & Partner www.stoploss.ch The System Trader No
Strategic Trading Systems, Inc. www.traderstech.net Checkmate Trading System; others Yes
Strategy Runner www.strategyrunner.com Not provided Yes
StrategyBot Inc. www.strategybot.com StrategyBot No
Striker Securities striker.com Striker Securities trading systems Yes
STT, LLC www.spb.addr.com Swinger-DT; Swinger-EOD Yes
SureFireThing.com www.surefirething.com Camarilla Equation Yes
Synthesis Bank www.tradingfloor.com tradingfloor Yes
T.B.S.P. Inc. www.TBSP.com The Right Time Yes
Taurus Corporation www.TaurusCorp.com Grand Cayman System Yes
TECHRULES.COM www.techrules.com techrules.com Yes
THE 2000 CORPORATION advanced-stock-selection.com Strongest Stocks Yes
Tiger Software www.tigersoft.com Peerless Stock Market Timing/Tiger Power Rankeer Yes
TimingCharts.com www.TimingCharts.com COT Trend Trader Method No
Tom Jackson’s S&P Trader www.sp-daytrade.com S&P Daytrade Yes
Trade Risk Management, LLC www.traderiskmanagement.com traderiskmanagement.com No

TradeForYourLiving.com www.TradeForYourLiving.com MX2 Trading System Yes
TradeGuider Systems www.tradeguider.com TradeGuider Yes
TradeOMS www.tradeoms.com TradeOMS No
Tradeportal Securities tradeportal.com Tradematrix Yes
Traders International www.TradersInternational.com TIMES Yes
TradersCoach.com www.TradersCoach.com Applied Reality Trading (ART) Yes
Tradeself tradeself.com Tradeself Stock Picks Yes
TradeSystem, Inc. www.trade-system.com Aberration; Aztec; I-Master Yes
Trading System Lab www.tradingsystemlab.com Trading System Lab Yes
Trading Systems Analysis Group www.tsagroup.com Trade Oracle Yes
Trading Systems Design and Analysis www.mesasoftware.com BigBlue-2 (www.bigblue-2.com); others Yes
TradingPatterns.com www.tradingpatterns.com APS Automatic Pattern Search Yes
TradingSystemSolutions.com www.TradingSystemSolutions.Com The Complete Stock Trading Solution for New Trader Yes
Tradology www.tradology.com Trend Financial LLC Yes
Trendchannel www.trendchannel.com Trendchannel Yes
TrendPro, Inc. trendpro.com Floor Traders Tools 8.0; TrendPro 8.0 Yes
TripleTrend www.TripleTrend.com TripleTrend Yes
Trotter Trading Systems, Inc. www.trotter.com S&P 500 Direction Finder System Yes
TurnSignal Inc. www.turnsignal.biz TurnSignal and the TurnSignal indicators Yes
Turtle Talk Inc. www.russellsandsoriginalturtle.com/ The Turtle Trading System Yes
TurtleTrader.com www.turtletrader.com TurtleTrader.com Yes
VCIP Systems www.vcipsystems.com/VCIP_Index.htm VCIP-HP DayTrading System No
Wett-Johnson Trading, LLC www.currencytrading4profit.com Currency Trading System Trend/Reversal Yes
www.invest33.com www.invest33.com Amsterdam Method Yes
S&C
The information in Traders’ Resource is the most accurate at the time of posting and is subject to change. Because the vendors posting to Traders’ Resource are responsible for their own listing, Technical Analysis, Inc. declines any
and all liability for any representations made by the businesses and individuals listed. Nor can Technical Analysis, Inc. endorse any business or individual listed on Traders’ Resource. Technical Analysis, Inc. makes no warranties,
express or implied, as to the accuracy and reliability of claims herein. You agree to release Technical Analysis, Inc., together with its respective employees, agents, officers, directors and shareholders, from any and all liability and
obligations whatsoever in connection with or arising from your use of Traders’ Resource. If at any time you are not happy with the information posted to Traders’ Resource or object to any material within Traders’ Resource, your sole
remedy is to cease using it. This list is updated frequently. If you are aware of a business that should be listed, please email us at [email protected].

Stocks & Commodities V. 27:01 (73-83): Traders’ Tips by Technical Analysis, Inc.
TRADERS’ TIPS
Here is this month’s selection of Traders’Tips, contributed by various
developers of technical analysis software, all to help readers more
easily implement some of the strategies presented in this issue.
Internet users will also find these and most previous Traders’ Tips
at our website at www.Traders.com. To locate the various tips, use
our site’s search engine, or click on the Stocks & Commodities
magazine link on the left-hand side of our home page, then scroll
down to the “This month in S&C” heading in the middle section
and click on“Traders’ Tips.” For previously published Traders’ Tips,
visit the “Back issues archive” located at http://www.traders.com/
Documentation/FEEDbk_docs/backissues.html. Code can then be
copied and pasted into your program.
This month’s Traders’

Tips are based on Os-
car Cagigas’s article in
this issue, “The Megan
Ratio.”
Code in AmiBroker
Formula Language for Figure 1: TRADESTATION, MEGAN RATIO. Here are sample MEGAN ratio results shown in
the Megan ratio from a chart (lower right), RadarScreen (left), and Scanner (upper right). The chart shows changes
in MEGAN over time (red line). RadarScreen holds the S&P 100, sorted from high MEGAN
Cagigas’s article is in-
ratio to low. Scanner can search for high MEGAN ratios in a security list of any length.
cluded at the end of this
document. Additional code is presented here as contributed
by software developers. This article is for informational purposes. No type of trading
Readers will find our Traders’ Tips section in its entirety at or investment recommendation, advice, or strategy is being made,
the Stocks & Commodities website at www.Traders.com in given, or in any manner provided by TradeStation Securities or its
the Traders’ Tips area, from where the code can be copied affiliates.
and pasted into the appropriate program. In addition, the
code for each program is usually available at the respective Strategy: CagigasA
software company’s website. Thus, no retyping of code is
required for Internet users. inputs:
FundSize( 100000 ),
For subscribers, the AmiBroker code found in Cagigas’s BuyLookback( 20 ),
article can be copied and pasted into AmiBroker from the ExitLookBack( 10 ) ;
Subscriber Area at www.Traders.com. Login is required.
variables:
Quantity( 0 ),
MEGAN( 0 ) ;
Quantity = Floor( ( FundSize + NetProfit ) / Close /

100 ) * 100 ;
Buy Quantity shares next bar Highest( High,

F TRADESTATION: MEGAN BuyLookBack ) stop ;
Oscar Cagigas’s article in this issue, “The Megan Ratio,” Sell next bar at Lowest( Low, ExitLookBack ) stop ;
describes the Megan ratio (maximum exponential growth
MEGAN = MEGAN_Calc( FundSize ) ;
annualized). Cagigas compares Megan to the Sharpe ratio
and K-ratio, and suggests that Megan be utilized as a means
of comparing trading systems. He presents three strategies to Strategy: CagigasB
demonstrate Megan’s ability to improve the trader’s ability inputs:
FundSize( 100000 ),
to critique different strategy options. FastLength( 5 ),
We have implemented Megan in such a way that it can SlowLength( 20 ),
be used in charting, RadarScreen, and in the new Scanner ExitLookBack( 2 ) ;
(available in TradeStation 8.4). Cagigas’s three strategies variables:
are presented here. The code for running his first strategy in Quantity( 0 ),
MEGAN( 0 ) ;
RadarScreen and in the Scanner will be posted in the Trade
Station and EasyLanguage support forum. Quantity = Floor( ( FundSize + NetProfit ) / Close /
To download the EasyLanguage code for this study, go to the 100 ) * 100 ;
TradeStation and EasyLanguage Support Forum (https://www. if Average( Close, FastLength) > Average( Close,
tradestation.com/Discussions/forum.aspx?Forum_ID=213). SlowLength ) then
Buy Quantity shares next bar High stop ;
Search for the file “Megan.Eld.”
Stocks & Commodities V. 27:01 (48-53): Walk-Forward Testing by Jack L. Weinberg
TRADING SYSTEMS
System Comparison
Walk-Forward Testing
Here’s how you can use walk-forward
testing as a tool to compare trading
systems and gain insight into comparing
the future performance of systems.
by Jack L. Weinberg
R ecent advances in trading

system software and pow-
erful hardware capabilities
have provided a valuable
tool in evaluating the future viability of
trading systems — walk-forward test-
ing. This technique allows the trading
system developer to quickly assess if the
system will survive (or even prosper) in
the future. It does this by segmenting the
past into in-sample and out-of-sample
periods, and by optimizing each succes-
sive in-sample period to provide the val-
ues for the next out-of-sample period. In
addition, testing can be used to compare
one system with another. Walk-forward
testing provides a unique insight into
comparing the projected performance of
trading systems.
OPTIMIZATION AND ITS USES

Optimization of trading systems is used
in order to learn from the past those
values that are most likely to succeed in
the future. This is at the heart of technical
analysis — the concept that the future
(especially the near-term future) can be
predicted from the most recent past. The
values determined from optimization of
the most recent past should be used over
the next time period. In the context of
walk-forward testing, the most recent
past is called the in-sample period, and
the next time horizon is called the out-of-sample period. IMPORTANCE OF WALK-FORWARD TESTING
Optimization gives the trading system designer the ability Walk-forward testing is one of the most powerful tools
to determine the best values for variables in the in-sample developed to analyze the future performance of trading
period. When a trading system designer puts too many systems. For the purposes of this comparative analysis, it was
constraints into the system rules, then the system is curve- decided that the period over which the walk-forward optimi-
fitted. Walk-forward testing will show that if a system is zation would be conducted would be eight years, with the first
curve-fitted it may work for a period in the past, but it will fail four being the first in-sample period. Walk-forward testing
for periods that are out of sample. works as follows: the first in-sample period is selected, and
Period start Period end Dataset
1/1/2000 12/31/2003 In sample
1/1/2004 12/31/2004 Out of sample
1/1/2001 12/31/2004 In sample
1/1/2005 12/31/2005 Out of sample ■ Equity positions in round lots of 100 shares
1/1/2002 12/31/2005 In sample ■ Exit after 60 days or with a specific condition,
1/1/2006 12/31/2006 Out of sample whichever comes first.
1/1/2003 12/31/2006 In sample
1/1/2007 12/31/2007 Out of sample SYSTEM 1: PETERSON BAND SYSTEM
FIGURE 1: WALK-FORWARD TESTING. Here you see the dates
This trading system was described by STOCKS & COMMODI-
used for in-sample and out-of-sample data. TIES Staff Writer Dennis Peterson in a 2002 article titled
“Developing A Trading System” (for testing purposes, a
the variables selected are optimized for a chosen objective slightly simplified version was used here). The article con-
function. Those values are then used for the first out-of- tended that if price “tagged” the lower Bollinger band,
sample period. The process is then repeated for each set of in- confirmed by a certain level in the stochastic relative strength
sample and out-of-sample periods. index (RSI), then that would be a precursor to an increase in
In Figure 1, you can see the sets of dates that were used in price. In addition, we should avoid days with short candle-
this article’s analysis of the systems. With walk-forward stick bodies. This trading system is ideal for optimization via
testing we can stitch together the out-of-sample periods to get walk-forward testing due to the limits placed on each condi-
a clear idea of how the system would have performed over the tion. Walk-forward testing will also demonstrate that if there
total out of sample period from January 1, 2004, to December are too many conditions it would cause the system to be
31, 2007. overoptimized or curve-fitted.
OPTIMIZATION OBJECTIVES Indicator: Bollinger bands

A key question to be asked by any system designer when Bollinger bands adjust themselves to market conditions by
implementing optimization is: What is the objective to be being dependent on standard deviation of prices. During
optimized? Most system trading software will allow optimi- volatile periods, the bands widen (that is, they move away
zation of net profit, but typically, this is not what the real from the average), and during less volatile periods, the bands
world trader feels should be the most important objective. contract (they move closer to the average). Peterson further
To compare systems, I chose the K-ratio measure. Devel- enhanced the use of Bollinger bands in this system by basing
oped by Lars Kestner and detailed in an article in STOCKS & them not on the closing price, but on a weighted average price
COMMODITIES, this measure can be called one of the most (Price= (2*Close+High+Low)/4).
important measures of the performance of a system. The K-
ratio is a unitless measure of performance that can be com- Indicator: Stochastic RSI
pared across markets and time periods. Together, the Sharpe The stochastic RSI combines two popular technical analysis
ratio and K-ratio are the most important measures when indicators, stochastics and RSI. Whereas stochastics and RSI
evaluating trading strategy performance. For more informa- are based on price, stochastic RSI derives its values from the
tion on choosing an objective for a trading system, please see relative strength index (RSI); it is basically the stochastic
Quantitative Trading Systems by Howard Bandy. indicator applied to the RSI. The stochastic RSI is used to
show periods of great momentum, otherwise referred to as
Testing data overbought or oversold periods. The AmiBroker code can be
All tests here were performed on historical NASDAQ data found in the sidebar “Stochastic RSI.”
(2000–07) obtained from www.eoddata.com. The subset of
securities that were selected for the purposes of testing were the System entry and exit rules
current stock components of the NASDAQ 100 Index (NDX). The entry rules are displayed in Figure 2. If all the conditions
Money management rules used are true, then enter on the following day at the open.
In addition to the previously mentioned money manage-
■ Risk $10,000 per trade ment rules, exit on the following day at the open, if all of the
■ Maximum of 20 open positions at any one time;
starting equity of $200,000 Condition description AmiBroker code
Look for prices tagging
■ Commission and slippage of $50 per trade the lower Bollinger band LoBand > Low AND High > LoBand;
Stochastic RSI should be
STOCHASTIC RSI FOR AMIBROKER above a threshold to ensure
some momentum is associated
function StochRsi(PeriodRSI) { with the push up StochRSI(PeriodRSI) > RSIUpLimit;
return 100*(( RSI( PeriodRSI) - LLV( RSI( PeriodRSI ) , PeriodRSI ) )
/ ( (HHV( RSI( PeriodRSI) , PeriodRSI ) ) - LLV(RSI( PeriodRSI ), The (close-open)/high-low) > 0.2
PeriodRSI ) )); to avoid days that have short
} candlestick bodies (Close-Open)/(High-Low) > 0.2;
FIGURE 2: ENTRY CONDITIONS FOR PETERSON BAND SYSTEM

DISH - Daily 4/14/2003 Open 28.01, Hi 28.47, Lo 27.88, Close 28.47 (3.0%) BBTop(18,1.8) = 30.17, BBBot(18,1.8) = 28.14, Close = 28.47
34.0
33.7819
33.0
32.0
31.62
31.0 exponential moving average shortens,
30.0 and the average is more responsive to
29.0
changes in price. The code that was used
28.2145
to calculate VIDYA can be seen in sidebar
28.0
“Calculation of VIDYA.” In that article,
27.0 Chande suggests using bands 1% above
26.0 and below VIDYA. This amount was
25.0 dynamically optimized in the code with
the variable VidyaDelta.
March April May
DISH - StochRsi(10) = 25.33 100.0
Indicator: Dynamic 75 80.0
60.0
momentum index 50
40.0
25
20.0
Continuing to make system indicators
March April May
8.45718
respond better to underlying price ac-
FIGURE 3: TYPICAL SYSTEM ENTRY FOR THE PETERSON BAND SYSTEM. Bollinger bands (in blue) are
tion, Chande first described the dy-
based on the weighted average price. The stochastic RSI, with the upper and lower limits set at 75 and 25, namic momentum index in a 1993
respectively, is displayed in the subchart. STOCKS & COMMODITIES article, “Sto-
chastic RSI And Dynamic Momentum
following are true: Index” (written with Stanley Kroll).
■ If the low is less than the upper Bollinger band, and Essentially, the DMI is a variable-length RSI, with its maxi-
■ StockRSI(Period RSI) < 100 - RSIUplimit mum and minimum number of days constrained. Chande
uses a minimum cutoff of 5 and a maximum cutoff of 30 in
All variable values (for example, PeriodRSI, RSIUpLimit, order to calculate this indicator. The code used to calculate
and so on) were determined via optimization in the in- DMI is shown in the sidebar “Calculation of DMI.”
sample period.
System entry and exit rules
Typical system entry If all the conditions displayed in Figure 4 are met, enter on the
In Figure 3, we see a chart of Dish Network Corp. (DISH) from following day at the open. In addition to the previously
March to May 2003. Bollinger bands (in blue) are based on mentioned money management rules, exit on the following
the weighted average price. Below the main chart is a chart of day at the open if all of the following are true:
the stochastic RSI, with the upper and lower limits set at 75
■ If the low is less than the upper VIDYA band, and
and 25, respectively (determined from optimization over that
time period). The green up arrow and the red down arrow are ■ DMI < 100 - uplimit
typical entry and exit points for the system.
All variable values were determined through optimization.
SYSTEM 2: CHANDE
BAND SYSTEM CALCULATION OF VIDYA
This system was first described in a 1995 STOCKS & COM-
MODITIES article, “Identifying Powerful Breakouts Early,” by function Vidya(Length, smooth) {
Tushar Chande. Chande’s contention was that “combining a Change = C - Ref( C, -1 );
variable length moving average (VIDYA) and a variable length //calculation of the day to day change in the Close
momentum oscillator (dynamic momentum index, or DMI) SU = Sum( C - Ref( C, -1 ) > 0, Length );
may help identify high-powered breakouts early.” Clearly, this //the sum of the Up days
was a unique attempt to harness the volatility of the security SD = Sum( C - Ref( C, -1 ) < 0, Length );
itself in trading. One of Chande’s key techniques was to make //the sum of the down days
indicators respond to market volatility by changing the period- CMO = 100 * ( SU - SD ) / ( SU + SD );
icity of the indicator. //calculation of the Chande Momentum Index
AbsCMO=( abs(CMO) )/100;
Indicator: VIDYA bands //derivation of the exponential length factor
VIDYA is a type of adaptive moving average that adapts to SC=2/(Smooth+1);
prices to show significant changes in volatility. What is //further derivation of the exponential length factor
unusual about VIDYA is that it goes sideways when there is
return AMA( C, SC*AbsCMO );
greater volatility. It is a variable-length exponential moving
//calculation of the Adaptive Moving Average –driven by CMO
average that is indexed to volatility or momentum. Therefore,
}
when market volatility increases, the effective length of the

CALCULATION OF DMI
function VarPeriodRSI( array, periods )
{
Chg = array - Ref( array, -1 );
Typical system entry pc = Max( Chg, 0 ); Jack L. Weinberg is a private trader
In Figure 5, we see a chart of Whole nc = Max( -Chg, 0 ); who has been active in trading stocks
Foods Market (WFMI) over a period pa = AMA( pc, 1/periods ); and commodities for more than 25
from January to May 2003. We see the na = AMA( nc, 1/periods ); years.
two parallel bands of VIDYA, with the return 100 * pa / ( pa + na );
parameters that have been determined } SUGGESTED READING
by optimization for that time period: STDa = MA(StDev(Close, 5), 10); Bandy, Howard B. [2007]. Quantita-
1.2 for VidyaDelta, 8 for VidyaLen, Vi = StDev(Close, 5)/STDa; tive Trading Systems, Blue Owl
and 20 for VidyaSmooth. Underneath TD = int(14/Vi); Press.
we see a chart for the DMI (in red) over TD= IIf(TD >30, 30, IIf(TD < 5, 5, TD)); Chande, Tushar S. [1992]. “Adapt-
that period, with the upper limit of 75 DMI = VarPeriodRSI(Close,TD); ing Moving Averages To Market
(again determined by optimization). Volatility,” Technical Analysis of
The green up arrow and the red down arrow are typical entry and STOCKS & COMMODITIES, Volume 10: March.
exit points for the system. _____, and Stanley Kroll [1993]. “Stochastic RSI And Dy-
As you can see from Figure 6, both systems performed well namic Momentum Index,” Technical Analysis of STOCKS
in the out-of-sample testing. Of special significance is the & COMMODITIES, Volume 11: May.
consistently high net profit and K-ratio in all _____ [1995]. “Identifying Powerful Breakouts Early,” Tech-
out-of-sample years, which would indicate a nical Analysis of STOCKS & COMMODITIES, Volume 13:
robust system. Overall, the Peterson band system October.
took 89 trades, with 47 of them being winners
(52.8%), while the Chande band system took Condition description AmiBroker code
225 trades, of which 127 (56.4%) were winners. Close above the upper band… conservative Close >= UpBand AND (Ref(Low,-1)
Although the net profit of the Peterson band approach… low above the upper band >= Ref(UpBand,-1));
system was lower, the exposure to the market was
also lower, causing a higher risk-adjusted return. We can use the DMI to measure momentum, averring
that the DMI rise above… to confirm the presence of
Out-of-sample equity line comparison a powerful breakout DMI > UpLimit;
Figure 7 shows the out-of-sample equity for FIGURE 4: ENTRY CONDITIONS FOR THE CHANDE BAND SYSTEM
each of the two systems under consideration,
along with a linear trendline. Of significance
to note is that not only is the Chande band WFMI - Daily 3/18/2003 Open 27.05, Hi 27.67, Lo 27.035, Close 27.64 (1.9%) UpperBand(1.2,8,20) = 25.28, LowerBand(1.2,8,20) = 24.68
system producing a higher net profit, it is 31.0
doing so consistently witnessed by the higher
correlation coefficient on that linear trendline 30.0
R2 = 0.9461 on the Chande system. This high
R2 value gives us greater confidence that the 29.0
system will continue to do well even in the
future.
28.0
28.0172
GAINING INSIGHT 27.3528

27.11527.0
This article has shown that walk-forward test-
ing can produce an out-of-sample equity line,
which is an excellent proxy for how the system 26.0
would actually have done during that time pe-
riod. Analyzing that out-of-sample equity line 25.0
can give the trading system designer an excel-
lent insight into the future performance of the March April May
system. WFMI - DynamicMomentumIndex(15) = 80.60
80.0
75
70.0
60.0
50 50.0
Analyzing that out- 40.0
30.0
of-sample equity line 29.4884
25
can give the system FIGURE 5: TYPICAL SYSTEM ENTRY FOR THE CHANDE BAND SYSTEM. Here you see two parallel bands
designer insight. of VIDYA with the following parameters: 1.2 for VidyaDelta, 8 for VidyaLen, and 20 for VidyaSmooth. The
DMI with the upper limit of 75 is displayed in the subchart.

Kestner, Lars [2003]. Quantitative Trading Strategies, Peterson, Dennis D. [2002]. “Developing A Trading Sys-
McGraw-Hill. tem,” Technical Analysis of STOCKS & COMMODITIES,
_____ [1996]. “Measuring System Performance,” Technical Volume 20: August.
Analysis of STOCKS & COMMODITIES, Volume 14: March. ‡AmiBroker ‡www.eoddata.com
Pardo, Robert [1991]. “Walking Forward Can Keep A Trad-
ing Model One Step Ahead,” Futures, July. S&C
% Of Winners
Peterson Band System:
# Of Winners
Max. Trade
Drawdown
# Trades
Exposure
K-Ratio
Sharpe
Payoff
Begin
Net %
Profit
Profit
Ratio
Ratio
RAR
End
Net
%
1/1/2004 12/31/2004 18,348 9.17 12.76 72.08 -3396 8.99 1.32 0.1568 13 8 61.54
1/1/2005 12/31/2005 16,103 8.05 32.93 24.73 -2601 2.97 0.39 0.0871 33 15 45.45
1/1/2006 12/31/2006 5,786 2.89 28.51 10.29 -2998 1.29 0.13 0.0601 32 17 53.13
1/1/2007 12/31/2007 28,948 14.47 9.79 149.16 -2720 7.35 0.97 0.0643 11 7 63.64
% Of Winners
Chande Band System:
# Of Winners
Max. Trade
Drawdown
# Trades
Exposure
K-Ratio
Sharpe
Payoff
Begin
Net %
Profit
Profit
Ratio
Ratio
RAR
End
Net
1/1/2004 12/31/2004 53,510 26.76 56.6 47.42 -4125 2.36 0.55 0.0390 64 37 57.81
1/1/2005 12/31/2005 22,701 11.35 56.99 20.15 -4134 1.7 0.52 0.1000 65 37 56.92
1/1/2006 12/31/2006 27,624 13.81 48.07 29.16 -2662 2.11 0.41 0.0953 56 29 51.79
1/1/2007 12/31/2007 11,056 5.53 35.05 15.91 -3183 1.45 0.33 0.1080 40 24 60.00
FIGURE 6: WALK-FORWARD TESTING: OUT-OF-SAMPLE RESULTS. An indication of the robustness of both

systems can be seen from the high percentage of winners in the last out-of-sample year, 2007. Overall, the
Peterson band system took 89 trades, with 47 of them being winners (52.8%). The Chande band system took 225
trades, of which 127 (56.4%) were winners. Although the net profit of the Peterson band system was lower, the
exposure to the market was also lower, causing a higher risk-adjusted return. In all, the K-ratio values are superior
for the Chande band system, indicating a smoother, more consistent out-of-sample equity line.
400000
2
R = 0.9461
350000
300000
2
R = 0.9273
250000
200000
150000
01/02/2004
03/02/2004
05/02/2004
07/02/2004
09/02/2004
11/02/2004
01/02/2005
03/02/2005
05/02/2005
07/02/2005
09/02/2005
11/02/2005
01/02/2006
03/02/2006
05/02/2006
07/02/2006
09/02/2006
11/02/2006
01/02/2007
03/02/2007
05/02/2007
07/02/2007
09/02/2007
11/02/2007
Peterson Band Chande Band Linear Trend (Chande Band) Linear Trend (Peterson Band)
FIGURE 7: OUT-OF-SAMPLE EQUITY LINE COMPARISON. The Chande band system produces a higher net
profit consistently witnessed by the higher correlation coefficient on that linear trendline R2 = 0.9461 on the
Chande system. This high R2 value gives us greater confidence that the system will continue to do well, even
in the future.

Stocks & Commodities V. 27:01 (70-71): Website Review: eMiniZ.com by Jayanthi Gopalakrishnan
EMINIZ.COM
Back in September 2006, we reviewed
www.eminiz.com, a website developed
by John F. Ehlers. Since then there have
been several additions and enhance-
ments that make it worth looking at
again.
When you visit the site, the first thing
that will catch your eye are four gauges.
For those of you who are familiar with
John Ehlers’s work on corona charts (an
article on the subject was published in
the November 2008 issue of S&C) will
recognize the terms such as cycle pe-
riod, cycle swing position, cycle S/N
ratio, and the adaptive trend vigor. If
you’re not familiar with corona charts,
the details about what the gauges indi-
cate are displayed prominently on the
home page, and you also have access to FIGURE 1: DOMINANT CYCLE PERIOD. Here you see a price chart with the trend swing indicator
overlaid. The corona chart is displayed in the subchart displaying the dominant cycle.
other features that will help you under-
stand them better. You can also down-
load EasyLanguage indicator code for overview of what’s behind this advanced is look at the daily updates and the
corona charts. software architecture. There is, on the performance results. Those who are not
website, a description of the eight trad- subscribers will get an idea of what this
FEATURES ing systems designed by John Ehlers. site offers by viewing the 30-day de-
This site provides trading signals for Being familiar with Ehlers’s work, I layed alerts and updates. But subscrib-
five mini index futures, namely the know it is very “techy.” The software ers will receive end-of-day updates af-
emini, NASDAQ 100 index, emini Russell was developed by eMINIZ cofounder ter the markets close as well as alerts for
2000 index, emini Dow 30 index, and Ric Way, another great technical mind. the next day’s open (Figure 2). In addi-
the emini S&P MidCap index. From the There is a technical paper you can down- tion to the daily alerts, you can also see
dropdown menu you can select the in- load that includes a discussion of the the details of each trade along with
dex you are interested in. technology behind eMINIZ. performance charts. One feature I liked
Going back to the gauges, if you want To truly understand how sophisti- about the performance bar charts was
to find out more about them, try clicking cated this system is, all you have to do that if you hover your mouse over a bar,
on the “Corona charts” part of the menu
to the left of the gauges. From there you
can select the topic of your interest.
Select the dominant cycle period and
select the emini NASDAQ 100 index.
What you will see is a price chart with
the trend swing indicator overlaid as
well as the corona chart that reflects the
dominant cycle period (Figure 1). You
can do the same for the cycle swing
position, cycle signal/noise ratio, or the
adaptive trend vigor, or you can look at
them together to determine where prices
are relative to the cycle. This, of course,
will help make your trading decisions.
The objective of eminiz.com is to
predict short-term swings in the five
indexes I mentioned earlier. It does this
through proprietary computer algo- FIGURE 2: UPDATES AND ALERTS. This is an example of what you can expect to get at the end of
rithms. I would like to mention here that every trading day. You see the current positions as of the close of the trading day as well as alerts for
even though it is proprietary, you get an the next day’s open.

Stocks & Commodities V. 27:01 (70-71): Website Review: eMiniZ.com by Jayanthi Gopalakrishnan
you will see the details of that specific out-of-sample performance. This avoids This is a site worth a visit regardless
trade. having to rely on backtested results. of where you are on the technical analy-
Other features include performance There is also a very active user forum sis curve. Beginners will get an idea of
reports, equity growth charts, profit trail where you can get any questions an- where they should be heading, and more
charts, and monthly profit/loss charts. swered about the product. I found some advanced technical analysts will find
One feature I would like to emphasize interesting discussions on some of the the highly technical aspect impressive.
here is the Monte Carlo simulation. This methods used to develop the eMINIZ As far as the site design goes, I thought
gives you, in addition to a realistic look system. it was easy to navigate, and the displays
at the performance of the system, an What does it cost to subscribe to this were simple, clear, and straightforward.
idea of how much thought and work service? The rates vary, depending on I thought it was a very effective way to
went into developing the system. What how many indexes you wish to receive communicate something that is diffi-
sets eMINIZ apart from any other trad- alerts and updates for and whether you cult and complicated.
ing system, besides the fact that it was wish to pay monthly or every four —Jayanthi Gopalakrishnan, Editor
created by two great technical minds, is months. The rates are clearly displayed
that it reevaluates multiple trading strat- on the website and you will find the
egies over multiple time frames and range anywhere from less than $100 to
selects the optimal strategy based on less than $200 per month. S&C

Stocks & Commodities V. 27:01 (44-47): Working Money: Managing Your 10-Bagger Portfolio by Thomas Maskell
Yr.
2nd 4.00
~$
Yr.
1st 00
$2
Managing Your 10-Bagger Portfolio

It’s a lofty goal to go for a 10-fold increase in Buying stock is always about the future, which is always
ANDREW DUTKOWSKY
the price of a stock, but is it realistic? about uncertainty, which takes us into the realm of probabili-
ties and statistics. It is here that we can predict the best, most
likely, and worst outcomes of our 10-bag strategy.
by Thomas Maskell When it comes to 10-baggers, we strive for the best. The
best can make a poor man rich. Conversely, we should avoid
O ver the past seven articles we have defined, analyzed,

purchased, and sold the 10-bagger. All our work went
toward one end: to reap the rewards of a 10-fold increase in the
stock’s price. It is a lofty goal, but is it realistic?
the worst, which will make us poorer, or even worse, fearful.
More often than not, we will achieve neither best nor worst.
We will end up somewhere in the middle, but where is that
middle? In this, our final 10-bagger installment, we will
answer that question.
LOOKING AHEAD LINKING PAST TO FUTURE

To answer that question, we need to go beyond the certainty All stock buys look toward the future, and all stock strategies are
of hindsight and venture into the probabilities of foresight. extrapolations. We hope what was will be, which is a leap of
TRADER’S NOTEBOOK
We have reached the end of our
journey into the world of 10-
faith. What gives us the confidence to make that leap is a rigorous baggers. The only thing left to do
analysis of potential outcomes. While there are many potential is go out and find them.
outcomes, our analysis focuses on just three: the best, the worst,
and the most likely.
The best outcome is that all our picks will be 10-baggers.
The worst is that all our picks will be losers from the point of sumptions. First, the best-case scenario:
purchase. And the most likely is a reasonable portfolio of the Best case: All of our picks are 10-baggers yielding a
best, the worst, and the in-betweens. It is this most likely typical annual return of 116%.
outcome that will define the potential success or failure of our
10-bag strategy. 100% of the portfolio x 116 annual return (1.00 x 116%)
We can build this most likely outcome by combining = 116% annual portfolio return
what we know about 10-baggers with a weighted average
analysis. However, it is important to note that what we know In the best case, if every one of our stock picks were 10-
about 10-baggers is situational; it is different for different baggers, our annual return on our total portfolio would be
types of markets. In a bull market there will be more 10- 116%, well above our minimum goal of 12%.
baggers, more advancers, and fewer decliners. A bearish Let’s look at the worst case. In this case, we anticipate that
market will provide a very different ratio of 10-baggers, the stock will decline from our buy point and trigger a moving
advancers, and decliners. The example given here is based average convergence/divergence (MACD) sell signal. Our
on the market conditions that were in effect at the time the analysis of the MACD sell signal indicated that the 10-
study began (2006). bagger’s price declines as much as 40% before it is triggered.
While our goal is to catch a string of 10-baggers, at a We’ll use this 40% for our analysis:
minimum, we want to do no harm. Since the stock market Worst case: All of our picks will immediately
provides, over the long term, a 12% return, our 10-bag decline yielding an annual return of -40%.
portfolio should offer an average return that is between the
market average (12%) and the typical annual return of a 10- 100% of the portfolio x -40 annual return (1.00 x -40%)
bagger (116%). = -40% annual portfolio return
THE BEST AND WORST

MOST LIKELY SCENARIO
We start our analysis with some reasonable assumptions:
The best and worst cases are easy to calculate, but as Figure
1 The market will continue to perform as it did when 1 illustrates, determining the most likely is much more
we gathered our data. In this case, it was a flat complex. The complexity arises from the various levels of
market. performance observed in the stock market ranging from the
2 We will sell when the stock reaches its 10-bagger typical 10-bagger all the way down to bankruptcy. So for
point. illustrative purposes, we need to limit our analysis in order to
make the calculation manageable.
3 Prior to its 10-bagger point, we will sell when it trips We’ll begin by creating a table (Figure 2), which is based
a MACD sell signal. on some of the data that we uncovered in our previous articles
4 All our signals will be based on monthly chart data. on the 10-bagger.
Any or all of these assump- Stock Chart for a Typical 10-Bagger

tions can be modified to fit your Price
14
trading strategy. For instance,
you could choose to buy using Best
12
the monthly chart but sell using 10-Bags, 116%
10
the weekly chart. You could
choose to sell at the 10-bag point 8-Bags, 100% 8
Dollars
only after the point has been 6

crossed in the hopes of capital- 5-Bags, 75%
izing on a 20-, 30-, or 50-bag- 3-Bags, 50% 4
ger. These are assumptions, 2
2-Bags, 25%
which means, with the excep-
tion of #1, they are strategic Loss, -40%
elements that you control. Launch 1st Yr. 2nd Yr. 3rd Yr.
~$1.00 ~$2.00 ~$4.00 ~$10.00 Worst
Let’s develop a weighted 10-
bag portfolio using these as- FIGURE 1: THE TYPICAL 10-BAGGER
Potential Percent
Category Annual Return
Bag Status of Market
1 +100% 10.1 4.2
2 75 to 100% 8.0 4.1
Using this data, we can construct another table based on a 3 50 to 75% 5.4 4.2
three-year scenario. Here we will assume that all the stocks 4 25 to 50% 3.4 8.4
in the portfolio trigger their sell point. This would not 5 0 to 25% 2.0 29.1
normally be the case, but it allows us to make the calculation. 6 Less than 0% None 50.0
We also have to choose a specific return rate for each FIGURE 2: TYPICAL PRICE HISTORIES IN A FLAT MARKET
category. We will choose 116% for the 10-baggers and the
high end of the range for the others. We can now build the Maximum Annual
table shown in Figure 3. Buy Annual Sell Price Percent of
Price in Weighted
Price Return % @ MACD Portfolio
The weighted return in the table is a three-year return 3 Years Return
assuming a $1.00 investment. To convert it to an annual (A) X
return will require a compound interest rate table. Consulting (A) (B) {(B)/100}
the table yields an annual return rate of about 17%. Clearly, 1.00 116 10.10 10.10 4.2 0.42
our 10-bagger portfolio has done no harm. It has surpassed 1.00 100 8.00 5.04 4.1 0.21
the long-term market return (12%) by 42%. And that was 1.00 75 5.40 3.24 4.2 0.14
accomplished based on a random screening of the market. 1.00 50 3.40 2.04 8.4 0.17
But what if our stock picks were not random? What if we 1.00 25 2.00 1.20 29.1 0.35
applied what we have learned about 10-baggers to increase 1.00 -40 1.00 0.60 50.0 0.30
the accuracy of our picks? Total Return 1.59
Figure 4 illustrates a 10-bagger portfolio that is not based FIGURE 3: THREE-YEAR RETURN ON A RANDOM 10-BAG PORTFOLIO (FLAT
on a random distribution of stocks. It is based on a select MARKET)
distribution that increases the number of high-end performers
and reduces the number of low-end losers. This is achieved by
Maximum Annual
applying some of the attributes of 10-baggers that we have Buy Annual Sell Price Percent of
Price in Weighted
discovered in our research. Attributes include price/sales be- Price Return % @ MACD Portfolio
3 Years Return
low 1.00, gross margin above 35%, and volume confirmation
(A) X
at launch. By combining these and other attributes we can (A) (B) {(B)/100}
reasonably expect our portfolio to outperform the random one.
1.00 116 10.10 10.10 9.2 0.93
Our annual return on this portfolio is 32%, almost three
1.00 100 8.00 5.04 9.1 0.46
times the long-term rate for the market. This entire portfolio
1.00 75 5.40 3.24 4.2 0.14
will increase 10-fold in less than eight and a half years. Even
1.00 50 3.40 2.04 8.4 0.17
with 40% of the picks being losers, a $5,000 investment based
1.00 25 2.00 1.20 29.1 0.35
on Figure 4 will make you a millionaire in just over 19 years.
1.00 -40 1.00 0.60 40.0 0.24
The data in Figure 4 is just one of many possibilities. For Total Return 2.29
instance, what would your annual return be if you limited FIGURE 4: THREE-YEAR RETURN ON A SELECT 10-BAG PORTFOLIO (FLAT
your downside to -30%? What if your selection process was MARKET)
conservative enough to eliminate any immediate downturn
stocks? The analysis can be made as complex or as simple as
you want and your skills as a stock picker allows. This is all SUGGESTED READING
about you: your strategy, your analysis, and your picks. Maskell, Thomas [2008]. The Complete Guide To Investing
During Retirement, Adams Media.
AIM FOR THE STARS _____ [2008].“Selling The 10-Bagger,” Technical Analysis
We have reached the end of our journey into the world of 10- of STOCKS & COMMODITIES, Volume 26: September.
baggers. We have discovered what they look like; how they _____ [2008].“The 10-Bagger Prelaunch,” Technical Analy-
develop; why they develop; and what they offer. The only sis of STOCKS & COMMODITIES, Volume 26: July.
thing left to do is go out and find them. As my analysis here _____ [2008].“Speculating With The 10-Bagger,” Technical
illustrates, by aiming for the stars you may only hit the moon, Analysis of STOCKS & COMMODITIES, Volume 26: April.
but the view from the moon is pretty spectacular. _____ [2007]. “Trading The 10-Bagger,” Working-
Money.com, December 4.
Thomas Maskell is a retired engineer and MBA with a modest _____ [2007]. “Search For The 10-Bagger Fundamentals,”
nest egg and a fundamental curiosity about the workings of the Working-Money.com, July 3.
stock market. Direct comments and questions on this article to _____ [2006]. “The Anatomy Of The 10-Bagger,” Working-
[email protected]. Include “10-bagger” in the subject line Money.com, September 5.
to avoid the tyranny of the delete key. Be sure to look for his The _____ [2006]. “The Search For The 10-Bagger Begins,”
Complete Guide To Investing During Retirement, on sale now. Working-Money.com, March 23.
Or visit his website at http://hometown.aol.com/retireestocks. ‡BigCharts.com S&C

TASC Magazine January 2009

Uploaded by

Copyright:

Available Formats

TASC Magazine January 2009

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

TASC Magazine January 2009

Uploaded by

Copyright:

Available Formats

Stocks & Commodities V.

27:01 (98, 96, 97): At The Close by Adrienne Toghraie

Point (T) On The

S ome of the most important questions that a

POINT (T) ON THE LAFFER CURVE

Copyright (c) Technical Analysis Inc.

Copyright (c) Technical Analysis Inc.

Explore Your Options

Copyright (c) Technical Analysis Inc.

THE RSI MIRACLE

METAQUOTES SOFTWARE CORP.

existing in the universe. By reading a

Copyright (c) Technical Analysis Inc.

FUTURES FOR YOU

Major Trend Reversals And Short-Term Moves

Identifying And Timing

As displaying the Special K uses a calculation (see

not offered as the perfect Daily KST -50

ket and falling ones in a bear trend. As a 100

the most recent primary trend cycles are Daily KST

or sell signal is going to work. Note the horizontal line

S&P Financials (XLF) relative to S&P

see the same thing develop almost a year

0.20 IT’S SPECIAL

Copyright (c) Technical Analysis Inc.

Taking Your Trading Seriously

C ynthia, tell us how you got

Copyright (c) Technical Analysis Inc.

Copyright (c) Technical Analysis Inc.

Applying Hysteresis To Moving Averages

Join The Band

NEW TECHNIQUES $DJAAG (Dow Jones-AIG Agriculture Sub-Index) INDX ©StockCharts.com

4˚C (Tb). In this situation, we say that the Ta

Simple moving average C + ... + C t ± 1 + C t Σ Ci

EMA t = EMA t ± 1 + 2 C t ± EMA t ± 1

Mean absolute deviation

Average true range TR 1 = H 1 ± L 1

Moving band Moving average ± k * Volatility

Example: Bollinger Band BB(n,k)t = SMAt ± k * SDt

KC n,k,p = EMA t ± k * ATR t

Signals generation can be

MULTIPLE SETS MOVING BANDS

$SPX (S&P 500 Large Cap Index) INDX ©StockCharts.com

Copyright (c) Technical Analysis Inc.

ENTRY Signal: is above the wave 5, which is the case

Copyright (c) Technical Analysis Inc.

January 2009 • Volume 27, Number 1

Copyright (c) Technical Analysis Inc.

Copyright (c) Technical Analysis Inc.

Which System = Higher Return?

The MEGAN Ratio

THE GEOMETRIC MEAN

M any metrics can evaluate the performance

Copyright (c) Technical Analysis Inc.

Copyright (c) Technical Analysis Inc.

MEGAN = 84*log(1.03) = 2.48 ■ There are 252 trading days in a year.

#include <geom.afl> Disadvantages of the MEGAN ratio

30.0 Buy = Cond = H > Ref(HHV(H,20),-1);

20.0 but system 2 trades a lot more. The geo-