IFTA Journal 2021 Edition

A Professional Journal Published by The International Federation of Technical Analysts

21
Inside This Issue

4 Number-Based Sentiment Indicators

12 The Noise Trend
27 Determination of Time Target Zones for Price Targets of Classic Price Patterns
38 Emerging Currencies as Equity Earthquake Indicator
55 Coefficient Moving Average

“Weak leadership
can wreck the soundest
strategy; forceful
execution of even a
poor plan can often
bring victory.”
—Sun Tzu ISSN 2409-0271
 IFTA JOURNAL 2021 EDITION

Letter From the Editor

By Dr. Rolf Wetzer, CFTe, MFTA ......................................................................................................................3
EDITORIAL

Rolf Wetzer, Ph.D., CFTe, MFTA MFTA Papers

Editor and Chair of the Editorial Committee
[email protected]
Number-Based Sentiment Indicators
By Przemysław Smoliński, MFTA...................................................................................................................4
Regina Meani, CFTe
[email protected] The Noise Trend
Aurélia Gerber, MBA, CFA, MFTA By
 Mohamed M. Khedr, CFTe, MFTA........................................................................................................... 12
[email protected]
Determination of Time Target Zones for Price Targets of Classic Price Patterns
By
 Momen Atef El Shayal, CFTe, MFTA.......................................................................................................27
Send your queries about advertising Emerging Currencies as Equity Earthquake Indicator
information and rates to [email protected]
By
 Ron Albert Marcelino Acoba, CFTe, CMT, MFTA...................................................................................38
Coefficient Moving Average
By
 Mohamed Fawzy ElSayed Ali AbdAlla, CETA, CFTe, MFTA...................................................................55

Articles
The Ripples Effect: A Clearer View for Market Action and Price Patterns
By Mohamed Ashraf Mahfouz, CETA, CFTe, MFTA ....................................................................................63
Forecasting Major World Indices on Ichimoku
By Yukitoshi Higashino, MFTA ....................................................................................................................76

NAAIM Papers
Stock Trends and Trend-Based Trading Strategies—Backed by Large-Scale Back-Testing
Implemented by Automated Software System
By Kevin Luo.................................................................................................................................................. 84
Simple and Effective Market Timing With Tactical Asset Allocation
By
 Lewis A. Glenn, Ph.D............................................................................................................................... 90
Cover Photograph from iStock images
Book Review
How the Average Investor Can Use Technical Analysis for Stock Profits:
An In-Depth Work on Stock Market Technical Analysis, Mob Psychology,
and Fundamentals by James Dines
Reviewed by Regina Meani, CFTe.................................................................................................................97

Author Profiles................................................................................................................................................. 98

IFTA Staff......................................................................................................................................................... 100

IFTA Board of Directors................................................................................................................................ 100

IFTA Journal is published yearly by The International Federation of Technical Analysts, 1300 Piccard Dr., Suite LL 14 Rockville, MD 20850
USA. © 2020 The International Federation of Technical Analysts. All rights reserved. No part of this publication may be reproduced or
transmitted in any form or by any means, electronic or mechanical, including photocopying for public or private use, or by any information
storage or retrieval system, without prior permission of the publisher.

IFTA.ORG PAGE 1
 + 23% NET return 2020
Risk ‘moderate’

MAKE YOUR CAPITAL WORK FOR YOU

www.investui.com
Investui is a modern and inexpensive way to invest. The markets display patterns called
market effects or calendar effects. Many academic papers confirm the existence of market
effects. Professional investors and traders make handsome profits from investments in
these patterns. Investui allows retail investors to also profit from market effects. The
service monitors the markets and sends e-mail alerts when a market effect occurs. One
click in the e-mail converts the alert into a real position! The position is opened and closed
automatically. Clients can see all positions in real-time. Sensational.

Investui clearly stands out in Investing with market I set an alarm clock
a comparison with others. effects brings amazing for every Investui
returns. mail.

Niedenau 36 I 60325 Frankfurt-am-Main I +49 (0)69 3487 5071 I [email protected]

Alarmservice, kann nicht als Anlageberatung verstanden werden.
 IFTA JOURNAL 2021 EDITION

Letter From the Editor

By Dr. Rolf Wetzer, CFTe, MFTA

Dear IFTA Colleagues and Friends:

I hope that you and your family and friends came through the year healthy and safe. We
are living through strange times. Something unprecedented keeps us from doing what we
are used to. We shelter ourselves by staying at home, wearing masks, and restricting our
meetings with others. The same applies to IFTA. For the first time in our history, there will be
no in-person conference this year. We canceled Philadelphia, we will stay at home, and we will
not meet as usual. Also, for the first time, there will be an online conference.
But, despite the pandemic, there is still an IFTA Journal. Not business as usual, but
nevertheless as normal as possible. IFTA is international and so is this Journal. We have
gathered articles and papers from colleagues around the world.
In this issue, you will find five papers from our MFTA program. Starting with the first idea until the final grading
of the paper, our colleagues usually spend almost a year achieving their MFTA designation.
We also have two papers that NAAIM has kindly made available to us. NAAIM also had to cancel its annual
meeting, and hence, the winner of the Wagner Award will not be announced until later this year. We therefore are
publishing papers from previous years. Furthermore, we are more than happy about the long-
lasting cooperation with NAAIM. I want to thank NAAIM for its support and especially Susan
Truesdale for her kind cooperation.
We also have two articles from colleagues in Egypt and Japan. The authors were speakers at our …the greatest
last conference in Cairo. Since their presentations raised much interest, they both agreed to write
an article about their respective topics to make them available for a broader audience. resources for the
Last, but not least, the Journal traditionally closes with a book review from our Australian
colleague, Regina Meani. She is something like the faithful soul of the Journal. Always there, always Journal are our
willing to help and to support. Thank you for that.
Finally, I need to thank Aurélia Gerber for her help and especially Linda Bernetich. Without Linda colleagues from all
and her team, the Journal simply wouldn’t exist. Thanks for your patience and organizational skills.
I hope that you enjoy the Journal. over the world.
Best regards,
Dr. Rolf Wetzer, CFTe, MFTA

IFTA.ORG PAGE 3
 IFTA JOURNAL 2021 EDITION

Przemysław Smoliński, MFTA

Number-Based Sentiment Indicators Stanislawa Herbsta 2a/17

Warsaw, 02-784
By Przemysław Smoliński, MFTA Poland

+48 696 433 713

[email protected]

Abstract probability of a market bottom or peak. While a gradually increasing number of rallying or
Based on the above observation, commitment of traders
declining stocks in the portfolio serves as a confirmation of the trend’s strength and
Depending on the stock exchange, various types of data are indicators, named Synthetic Sentiment Indicators, were
contributes to its continuation, a decisive prevalence of either market side serves as a reliable
available to determine commitment of traders that can be used developed. Recent years have shown that they accurately
prognosis of a market extreme emerging and a reversal of the current long-term trend.
to help forecast future market behaviour. Such data may be determine the probability of occurrences of market peaks and
released directly by the exchange itself or be obtained through bottoms.
Based The high
on the above effectiveness
observation, commitmentof ofthe presented
traders indicators, indicators can
named Synthetic
various kinds of surveys of trader opinions and sentiment. be explained
Sentiment fromdeveloped.
Indicators, were a fundamental
Recent yearsperspective—based on the
have shown that they accurately
While the former type of data is not always available, given flowstheofprobability
determine capital in the economy
of occurrences andpeaks
of market in theandcapital
bottoms.market itself,
The high effectiveness
that such data is not published by all exchanges, the latter is of theand fromindicators
presented the perspective of technical
can be explained analysis
from a fundamental by using cycles
perspective—based on the
frequently burdened with different types of errors that arise flowsand a fractal
of capital marketand
in the economy structure.
in the capital market itself, and from the perspective of
from conducting the survey on a limited group of traders, technical analysis by using cycles and a fractal market structure.
the periodic nature of the surveys, etc. Help in solving the Technical Grounds
Technical Grounds
above problem came, rather unexpectedly, from emotionless The majority of main market peaks and bottoms can be
algorithms. An investment portfolio based on them turned out explained
The majority ofby a simultaneous
main superposition
market peaks and bottoms of extremes
can be explained from
by a simultaneous
to be a good indicator of market sentiment. severalofdifferent
superposition market
extremes from severalcycles.
differentIn addition,
market if we
cycles. In consider
addition, the the
if we consider
A detailed analysis of the above issue proved its solid fractal
fractal nature
nature of of the
the market, wheremarket,
the end ofwhere the end
the structures of thetimeframes
in various structures in
constitutes
theoretical foundation based on the theory of cycles and the variousoftimeframes
the beginning constitutes
a subsequent structure, the beginning
market extremes of a subsequent
can be forecasted on the basis of the
fractal nature of the market as well as capital flows in the structure,
current cycle phasesmarket
in variousextremes
timeframes.can be forecasted
The simultaneous on the basis
superposition of their peaks and
stock market and in the economy as a whole. More importantly, of the
bottoms current
results cycle phases
in the occurrence in various
of the main timeframes. The
market extremes.
however, it also has practical investment applications. This simultaneous superposition of their peaks and bottoms results
thesis presents the theoretical foundations of the described Figure 1. Interactions between
in the occurrence of thecycles  main market extremes.
issue and examples of its practical application.
Figure 1. Interactions between cycles
Acknowledgements Cycle long-term

I would like to start by thanking Paweł Małmyga for

exchanging many interesting investment ideas with me and for
the countless joint strategy backtests conducted on their basis. I Cycle mid-term

would also like to thank Emil Łobodziński for numerous hours of

inspiring conversations about the capital market, and my year-
long employer PKO BP Securities for providing me with working Cycle short-term

conditions that allow me to combine my work with my passion

for technical analysis.

Introduction Trend reversals and corrections

Every morning, the author of this thesis prepares an A superposition of short- and
mid-term cycles suggests the
algorithmic investment portfolio covering 140 of the biggest beginning and end of corrections

companies listed on the Warsaw Stock Exchange. After years

of monitoring signals and the general number of short and long
positions in the portfolio, a pattern has started to emerge that A superposition of short- and
proves helpful in determining market sentiment. As it turned mid-term cycles suggests the
beginning and end of corrections
out, a decisive prevalence in the number of equities on one side
of the market far increases the probability of a market bottom
or peak. While a gradually increasing number of rallying or Cycle Identification
declining stocks in the portfolio serves as a confirmation of the In classical technical analysis, one must first identify the
trend’s strength and contributes to its continuation, a decisive dominant cycles, determine their period, and find the most
prevalence of either market side serves as a reliable prognosis of recent highs and lows to predict respective market peaks and
a market extreme emerging and a reversal of the current long- bottoms. While it may theoretically seem uncomplicated, cycle
term trend. identification tends to be problematic in practice. To achieve this
goal, one can use visual methods based on manually measuring

PAGE 4 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

the distance between visible market extremes or, alternatively, better with respect to broad market indicators that aggregate
use more advanced methods, such as the Fourier analysis the current cycle position for all index constituents than with
developed by French mathematic John Baptiste Joseph Fourier, regard to individual equities. For an individual stock, the
the Phasing presented by J.M. Hurst, or the Maximum Entropy convergence of cycle phases, i.e., trends with the same direction
Spectral Analysis (MESA) described by John Ehlers. in various timeframes, may suggest a growing or declining
interest in the given equities on the part of consecutive groups
Figure 2. Traditional forecasting method using cycles of traders and translate into a continuation of the current
Forecast trend in the following sessions. Meanwhile, if a great majority
of equities consistently remains in a growth or a decline cycle
phase across all timeframes, the market can be considered
Forecast
widely overbought or oversold, respectively, which considerably
increases the probability of a market peak or bottom emerging.
While it is relatively easy to spot recurring structures on the As a result, one might say that, instead of forecasting a
charts during a long-time horizon, shortening the timeframe priori a market extreme, in the future, it is better to focus
usually results in increased chaos and decreased regularity on determining the probability of its occurrence now. The
in the occurrence of market extremes. An average 54-year above approach can help eliminate the need for a far-from-
Kondratieff Wave, 18-year Kuznets, and 9-year Juglar cycles or perfect manual forecasting of peaks and bottoms or for the
the Presidential Cycle on the U.S. market can serve as a case application of advanced mathematical methods which often do
in point here. Thanks to robust macroeconomic foundations, not seem fully adapted to the capital market and its constantly
these cycles are relatively easy to identify, and their impact on changing nature.
stock prices is clearly distinguishable. However, as the cycle
period is shortened, predicting market peaks and bottoms Macroeconomic Grounds
frequently becomes more difficult, and the results are prone to Although macroeconomics is not the focus of this thesis,
errors. Not only do the cycles frequently change their period and every good theory should have the most extensive theoretical
amplitude but they also tend to disappear and re-emerge after foundation possible. As a result, apart from the technical
some time. Their practical application is further complicated approach to the presented theory, we should also analyse its
by the superposition of several cycles of different periods and foundations based in the flows of capital in the economy and the
amplitudes. capital market.
A bull market, especially supported by the good condition
Figure 3. Probability of occurrences of peaks and of the economy as a whole, means a gradual capital circulation
bottoms based on cycles superposition
between equities. The capital steadily flows from the rallying,
overvalued stocks to less expensive ones (e.g., according to peer
valuation ratios) whose prices have so far been declining or
moving horizontally. The higher the number of stocks following
a strong uptrend, the fewer stocks that remain attractive in
terms of their price, which means that, following the profit-
taking from overvalued equities, the traders cannot reinvest
their capital in less expensive stocks. As a result, at some point,
In light of the above, we should consider applying a method traders begin to look for alternative means of capital allocation
of predicting cycle-based market extremes that is slightly which, combined with changing prospects of GDP growth or
different from the traditional method based on measuring the monetary policy, results in shifting the capital from the
the cycle period from its latest bottom or peak. If we make an equity market toward markets that are more attractive at that
assumption regarding the fractal nature of the market and the time, be it commodities, money, or debt instruments. Combined
resulting occurrence of the main market extremes at the points with the change in the economic cycle, a strong capital outflow
of simultaneous superposition of the end of structures in all contributes to a reversal to downtrend on the stock market.
time horizons (a short, medium, and long investment horizon Declining prices of equities result in a gradual increase in their
can be assumed for simplicity), we can simply restrict ourselves attractiveness, which in time causes another inflow of capital
to determining the current position of the price in respective from other areas of the capital market.
cycles of a specific period. As a result, instead of seeking specific The above explanation has an additional advantage over
cycle periods and measuring their distance from the last market the purely technical approach, because it also clarifies why a
extremes, we simply assume their existence and determine simultaneous superposition of cycles of the same phase and in
whether the price is currently in the growth or the decline different timeframes works better for forecasting extremes in
phase. Once the phases of the studied cycles coincide, the the broad market, using all the equities comprised by the given
probability of a market extreme emerging increases. index, than for individual equities.
As it turns out, the approach described above works even

IFTA.ORG PAGE 5
 IFTA JOURNAL 2021 EDITION

Figure 4. Money flow between sectors of the economy

analysis tools for trend determination. First, trend direction
must be established for every equity separately in the short,
medium, and long term. Next, the results should be summed
up for the whole market or respective indices. This can be
achieved through the application of many different technical
analysis tools, such as the simplest moving average. In this
thesis, a Trailing Stop was used, which is distant from the
stock price by a fixed ATR value that increases the more the
indicator is expected to serve the purpose of determining the
longer cycle phase.
Figure 5. Cycle identification based on price
Cycle identification based on price
Nasdaq, Trailing Stop Short-term, Trailing Stop Mid-term, Trailing Stop Long-term 5950

5900

5850

5800

5750

5700

5650

5600

5550

5500

5450

Short-term Trailing 5400

Stop identifies
Short-term Trailing Stop
5350

identifies short-term cycles

5300

short-term cycles 5250

5200

Mid-term Trailing
Materials and Methods Mid-term Trailing Stop
Stop identifies
identifies mid-term cycles
5150

5100

mid-term cycles
5050

5000

4950

Identifying Market Sentiment Long-term Trailing

Long-term Stop
Trailing
4900

4850

identifies long-term cycles

The abovementioned regularity can be used for creating Stop identifies 4800

long-term cycles 4750

Synthetic Sentiment Indicators to determine the scale of 4700

traders’ commitment and, at the same time, general market

4650

4600
28 5 12 19 27 3 9 17 23 30 6 13 21 27 6 13 20 27 3 10 17 24 1 8 15 22 30 5 12 19 26
December 2017 February March April May June

sentiment that prove helpful in establishing the strength Figure 5 Cycle identification based on price

of the current trend and the degree to which the market is Given that it is possible to use classical technical analysis
overbought or oversold, which is useful in forecasting the indicators
Given that itto determine
is possible cyclestechnical
to use classical instead of forecasting
analysis peakscycles
indicators to determine and
main turning points. bottoms manually,
instead of forecasting peaks the wholemanually,
and bottoms processthecan
wholebe automated
process withwith
can be automated
Given that neither the precise period of the cycles nor the relative ease for a large number of equities listed on the stock
relative ease for a large number of equities listed on the stock exchange or included in a
exact dates of peaks and bottoms are needed, and instead exchange or included in a specific index.
specific index.
are replaced with determining the current cycle phase in the
adopted investment horizons, the whole process can be largely Basic Interpretation
Basic Interpretation
automated based on properly prepared indicators applied to After aggregating the results for all equities included in the
the prices of all equities from the given index or market. Their analysed indices,
After aggregating we arrive
the results at a summary
for all equities ofanalysed
included in the the market inarrive
indices, we the at a
combination into an aggregate indicator determines the general form
summary ofofindicators illustrating
the market in the theillustrating
form of indicators currentthecycle
currentphase in various
cycle phase in various
market sentiment. timeframes that can be interpreted as a joint market sentiment
timeframes that can be interpreted as a joint market sentiment of short-, mid- and long-term
Assuming that the cycle amplitude is largely proportional of short-, mid-, and long-term traders. For most equities, the
traders. For most equities, the prevalence of the growth or decline phase of the analysed cycle
to its period, we can stipulate that the respective cycles are prevalence of the growth or decline phase of the analysed
represented in charts by trends in various timeframes. As a indicatesindicates
cycle the positive or negative
the attitudes
positive orofnegative
the traders towards the market.
attitudes of the traders
result, instead of determining the phase of the cycle based toward the market.
on the latest bottoms or peaks, we can use classical technical

PAGE 6 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Various
Figure 6. Various stages of market behaviour stages of market behaviour

DAX 13650
13600
13550
13500
13450
13400
13350
13300
13250
13200
13150
13100
13050
13000
12950
12900
12850
12800
12750
12700
12650
12600
12550
12500
12450
12400
12350
12300
12250
12200
12150
12100
12050
12000
11950
11900
11850

Overbought Balanced Market Oversold 11800

11750
11700
100
Short-term Sentiment
_DAX Sentiment Short-term

50

-50

-100

Mid-term Sentiment
_DAX Sentiment Mid-term

50

-50

Long-term Sentiment
_DAX Sentiment Long-term

50

-50

28 4 11 18 25 2 9 16 23 30 6 13 20 27 4 11 18 27 2 8 15 22 29 5 12 19 26 5 12 19 26 3 9 16
September October November December 2018 February March April

Figure 6 Various stages of market behaviour

A gradual increase in the number of equities in an uptrend or a downtrend implies a strong

IFTA.ORG PAGE 7
 IFTA JOURNAL 2021 EDITION

A gradual increase in the number of equities in an uptrend Backtests

or a downtrend implies a strong and healthy trend that should To determine the validity of the described theory, Synthetic
continue until a critical value is reached. Above it, a decisive Sentiment Indicators were tested on several markets from
majority of equities will be at the same cycle phase in all 2017–2019. The test results confirmed the effectiveness of the
the timeframes, which means that the whole market will be described method for DAX, STOXX 600, ASX 200, S&P 500, and
extremely overbought or oversold. If the market is overbought, Nasdaq Composite indices; it is likely that the positive effects of
the traders will not be able to reinvest their profits in cheaper its application would also apply with a high degree of probability
equities which will result in an outflow of capital in favour of in the case of other stock indices.
other, more attractive instruments, while stock indices will
reverse to a downtrend or at least start a strong correction. On Results
the other hand, if the market is extremely oversold, the majority The presented indicators can be applied both for the purpose
of equities will be sufficiently attractive in terms of price to of discretionary trading, in the scope of classical technical
induce the traders to turn their back on other segments of the analysis, and in a systemic approach as one of many components
capital market and the indices will begin long-term uptrends. of more complex investment strategies. In either case, the
The above description of the sentiment indicators is only basic application of the indicators focuses on determining the
basic. The indicators provide far greater analytic possibilities current dominant market trend and its further potential and
that make it possible to determine the current trend and signal determining the points in time where the probability of the
a trend weakening, an approaching correction or a long-term trend reversal is high.
trend reversal. Interpreting the Synthetic Sentiment Indicators involves
aggregating the results of the three separate indicators
Structure determining the sentiment of short-, mid-, and long-term
Synthetic Sentiment Indicators are based on four traders or aggregating the Long vs. Short Sentiment Indicator,
fundamental source indicators from which aggregate result which establishes the joint commitment of all trader groups in
indicators emerge: all investment horizons. In the former case, the basic signal of
• Short-term Cycle: determines the phase of the short-term a trend reversal involves all three indicators being overbought
cycle. or oversold at the same time. In the latter, a mutual position of
• Mid-term Cycle: determines the phase of the mid-term cycle. both lines of the Long vs. Short Sentiment Indicator serves as
• Long-term Cycle: determines the phase of the long-term a signal of trend continuation. Taking this into consideration,
cycle. the conducted tests were divided into mean reversion and trend
• Combined Cycles: determines the degree to which cycle following. An additional third test was also performed, taking
phases coincide in the short, mid, and long horizons. both categories into consideration.
The results of the abovementioned tests follow.
Each of the above indicators should be applied to every
individual equity included in the index, after which the results Mean Reversion
should be summed up, giving rise to the actual Synthetic In the case of the mean reversion test, the signal reversing the
Sentiment Indicators: current uptrend appears when all three indicators illustrating
• Short-term Sentiment: estimates the commitment degree of the sentiment of short-, mid-, and long-term traders are
short-term traders. overbought (above 50 points) at the same time. The situation
• Mid-term Sentiment: estimates the commitment degree of is reversed in the case of a sell signal that appears when all of
mid-term traders. these indicators are simultaneously oversold (below 50 points).
• Long-term Sentiment: estimates the commitment degree of In both cases, the position is closed automatically 10 sessions
long-term traders. from the signal’s appearance, which corresponds on average to a
• Long vs. Short Sentiment: estimates the degree of period of two weeks.
commitment consistency for all analysed trader groups on
the long or the short side of the market. Table 1. Mean reversion signals

Percent Avg. Profit /

The formulas for the above indicators can be found in the Index Trades Profitable Avg. Loss Profit
Appendix at the end of this thesis. ASX 200 7 86% 2.27 12.36%
As stipulated above, other classical technical analysis tools
DAX 12 67% 1.76 14.78%
can be used for constructing the aforementioned indicators,
NASDAQ 6 50% 2.97 8.98%
such as moving averages or oscillators. This thesis, however,
uses the Trailing Stop, which determines the current trend in S&P 500 7 71% 3.43 13.63%
a specified investment horizon in a simple and transparent STOXX 600 5 80% 1.91 10.94%
manner. As a result, it can easily be used for a large number of
equities, while the aggregate results clearly depict the general
market sentiment.

PAGE 8 IFTA.ORG
 Percent Avg. Profit /
Index Trades Profitable Avg. Loss Profit
ASX 200 7 86% 2.27 12.36%
DAX 12 67% 1.76 14.78% IFTA JOURNAL 2021 EDITION
NASDAQ 6 50% 2.97 8.98%
S&P 500 7 71% 3.43 13.63%
STOXX 600 5 80% 1.91 10.94%

following test when either line of the Long vs. Short Sentiment
Mean reversion signals on DAX
Indicator enters the overbought range (above 50 points). The
Figure 7. Mean reversion signals on DAX 3.3 TREND FOLLOWING WITH PROFIT-TAKING
position is renewed after the indicator declines beneath the
DAX 13700
13600
P 13500

P
Y Y
13400
13300
13200
13100
13000
overbought range, provided that the original conditions for
The last test conducted combines Trend Following signals with Mean Reversion by adding
opening the position are still met.
12900
Y 12800
12700
12600
P

profit-taking signals to the Trend Following test when either line of the Long vs Short
12500
P P 12400
Y 12300
Y O Y Y Y 12200
P Y O 12100
12000

Table
Sentiment3. Trend-following
enters the overboughtsignals with profit-taking
11900

Indicator range (above 50 points). The position is renewed

11800
P Y O 11700
11600
O 11500
11400
11300

after the indicator declines beneath the Percent

overbought range, provided that the
Avg. Profit / original
11200
11100
Y 11000
10900
10800

Index for opening

Trades
the position are Profitable Avg. Loss Profit
Y
10700

conditions still met.

10600
10500
10400
10300
O

ASX 200 40 55% 1.00 10.38%

10200
10100
_DAX Sentiment Short-term 100

50

0 Table
DAX 3 Trend following
32 signals with53%
profit-taking 1.54 33.10%
-50
Percent Avg. Profit /
Index Trades Profitable Avg. Loss Profit
_DAX Sentiment Mid-term
-100
100

50
NASDAQ
ASX 200
41 40 55%
54% 1.00
1.27 10.38%
30.65%
DAX 32 53% 1.54 33.10%
S&P 500 23 52% 1.51 27.73%
0

NASDAQ 41 54% 1.27 30.65%

-50

-100

S&P 500 23 52% 1.51 27.73%

STOXX 600 34 50% 1.26 8.54%
_DAX Sentiment Long-term 100

STOXX 600 34 50% 1.26 8.54%

50

-50

Dec 2017 Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2018 Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2019 Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2020

Figure 7 Mean Reversion Signals on DAX Figure 9. Trend-following signals with profit-taking on
S&P 500 Trend following signals with profit-taking on S&P 500
Trend Following
SP500 3300

3250

Y 3200

13
In the following test, trend-following signals are the result 3150

3100

3050

of the Long vs. Short Sentiment Indicator that illustrates the Y

O
P
Y

O
3000

2950

aggregate commitment of all trader groups in all investment

P P
Y
Y 2900
Y Y
P O 2850
P
O Y Y
Y 2800

horizons. A rally of the long line in excess of the short line

Y
O Y Y
Y P O
Y PO YP 2750
P O
Y P 2700
Y

generates
3.2 a buy signal and vice versa. In order to filter the
TREND FOLLOWING Y
Y
Y
O
2650

2600

signals in consolidations, the signals do not appear immediately

2550
P
2500

In the following test, Trend Following signals are the result of the Long vs Short
Y

when the lines cross but, instead, when the difference between
P O 2450
Y
2400

Sentiment Indicator that illustrates the aggregate commitment of all trader groups in all
them is at least five points. The position is closed immediately
2350

2300

after both linesA cross.

2250

investment horizons. rally of the Long line in excess of the Short line generates a buy signal
O

2200

_SP500 Sentiment Long vs Short 70

and vice versa. In order to filter the signals in consolidations, the signals do not appear
65
60

Table 2. Trend-following signals

55
50
45
40

immediately when the lines cross but, instead, when the difference between them is at least 35
30
25

Percent Avg. Profit / 20

15

five points. The position is closed immediately after both lines cross.
10

Index Trades Profitable Avg. Loss Profit

5
0
-5
Dec 2017 Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2018 Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2019 Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2020

Figure 9 Trend following signals with profit-taking on S&P 500

ASX 200
Table 17 signals
2 Trend following 29% 1.82 -5.38%
DAX 19Trades Percent
47% Avg. Profit /
1.97 20.46% Final Remarks on the Backtests
Index Profitable Avg. Loss Profit
ASX 200 17 29% 1.82 -5.38% The results presented above illustrate the effects of a basic
NASDAQ
DAX 20 19 47% 40% 1.97 2.26 20.46% 21.42%
NASDAQ 20 40% 2.26 21.42% application of the indicators to several main stock indices.
S&P 500
S&P 500 16 16 50% 50% 1.61 1.61 16.34% 16.34% 15
STOXX 600 17 41% 1.89 6.33%
It should be noted that the indicators’ parameters were not
STOXX 600 17 41% 1.89 6.33% optimised with respect to generating the highest profit possible;
instead, appropriate overbought and oversold levels were
Figure 8. Trend-following signals
Trend following signals oncomposite
on Nasdaq Nasdaq composite selected visually so that they better reflect the nature of the
Nasdaq

Y
9000
8900
8800
8700
8600
given market. It should also be borne in mind that the results of
the tests do not illustrate any tradeable investment strategies
8500
8400
8300
8200
8100
Y 8000

but only the raw verifiability of the indicators themselves, which

Y 7900
P P O
7800
O 7700
P Y Y
Y 7600
P 7500

should only serve as a basis for further and more advanced

O 7400
Y
7300
Y Y
P 7200
Y
7100
O
7000
O

works on the specific methods of their application.

Y Y 6900
P P O
6800
Y
6700
Y
O 6600
6500
Y
6400
6300
6200

Discussion
6100
6000
PY
5900
Y O 5800
OPO Y
5700
Y 5600
Y
5500
5400
5300

5200

5100

5000
Results Commentary
The test results presented above explicitly show the high
4900
O
4800

4700

effectiveness of the Synthetic Sentiment Indicators. A positive

_Nasdaq Sentiment Long vs Short
80
75
70
65
60

rate of return was achieved both in the case of mean reversion

55
50
45
40
35
30

signals, based on the simultaneously overbought or oversold

25
20
15
10
5

indicators in the short, medium, and long investment horizon,

0
-5
Dec 2017 Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2018 Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2019 Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2020

Figure 8 Trend following signals on Nasdaq composite

and the trend-following signals that are based on the Long vs.
Trend Following With Profit-Taking Short Sentiment Indicator, which determines the consistency
The last test conducted combines
14
trend-following signals with of the phases of all the cycles at the same time. In regard to the
mean reversion by adding profit-taking signals to the trend- mean reversion signals, a positive rate of return was generated

IFTA.ORG PAGE 9
 Indicator to create an oscillator that works better in the consolidation period while still
IFTA JOURNAL 2021 EDITION highlighting the points of the market being strongly overbought or oversold.

Long-Short
Figure 11. Long-Short Difference oscillator
Difference on DAX on DAX
oscillator
with respect to each of the tested indices, and the total profit DAX

13500

amounted to 60.69%. The test of the trend-following signals 13000

12500

returned slightly worse results; ASX 200 was the only index 12000

that generated a small loss, and the total profit for all the tested 11500

markets amounted to 59.17%.

11000

10500

As it turns out, a combination of both signal types generated 10000

the best results by far. Not only was a profit rather than a 9500

loss observed in the case of ASX 200, but practically all the _DAX Sentiment Difference
150

remaining indices generated a far better result than in either

100

50

test individually, achieving a total rate of return of 110.40%. -50

-100

-150

_DAX Sentiment Long vs Short 70

65

Discretionary Trading With Synthetic Sentiment

60
55
50
45
40

Indicators
35
30
25
20
15

Apart from using clear and unequivocal investment signals

10
5
0

y Jun Jul Aug Sep Oct Nov Dec 2017 Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2018 Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2019 Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2020

generated by the lines of respective indicators crossing or Figure 11 Long-Short Difference oscillator on DAX

the indicators entering overbought or oversold ranges, the Comparison With Other Broad Market Indicators
Synthetic Sentiment Indicators can also be applied to a more Finally, a comparison of the Synthetic Sentiment Indicators
subjective market assessment that utilizes less obvious nuances with other broad market indicators is recommended. The most
in the conduct of the indices. The mutual relations between popular among them are the Advance/Decline Line and the
respective indicators or their relative behaviour versus the 52 Week High/Low. In a great majority of cases, the Synthetic
exchange index can be used for this purpose. Sentiment Indicators are decisively better with respect to
Divergences
disclosing a strongly overbought and oversold market and
Divergences
The most straightforward example of using the discussed indicators are divergences. They creating divergences as well as generating trend reversal and
The most straightforward example of using the discussed continuation signals more accurately and quickly than classic
warn in advance of an increased probability of a correction start or a reversal of a long-term
indicators are divergences. They warn in advance of an increased broad market indicators that are commonly used.
trend. 19
probability of a correction start or a reversal of a long-term trend.
Figure 12. Synthetic Sentiment Indicators vs. A/D Line
Figure 10. Divergences on ASXon200
Divergences ASX 200 and 52 Week High/Low on S&P 500
ASX200
6400
Advance/Decline Line 
30000
6350

6300
25000
6250
52 Week High/Low Index 
6200 100
6150

6100 50

6050

6000 52 Week High/Low  0

5950 200
150
5900
100
5850
50
5800 0

5750 3050
3000
5700 2950
2900
5650
2850
5600 2800
2750
5550 2700
2650
5500
2600
5450 2550
2500
5400
2450

100 2400
_ASX200 Sentiment Cycle Short-term
2350

50 Short‐term Sentiment 
50
0
0

-50 -50

-100
Mid‐term Sentiment  -100

_ASX200 Sentiment Cycle Mid-term 50

50
0

-50
0
Long‐term Sentiment 
-50 50

0
_ASX200 Sentiment Cycle Long-term
-50
50
-100

Long vs Short Sentiment  60
0 50
40
30
20
-50 10
0
3 20 27 4 10 17 24 1 8 15 22 29 5 12 19 26 3 10 17 24 31 7 14 22 28 4 11 19 25 4 11 18 25 1 8 15 22 29 6 13 20 28 3 10 17 24 1 8 15 22 29 5 12 19 26 3 9
September October November December 2018 February March April May June July August September October November December 2019 February March
September October November December 2019 February March April May June July August Septembe

Figure 10 Divergences on ASX 200

Long-Short Difference Oscillator Conclusion

The discussed indicators are not the ultimate form of Although, as indicated in the introduction to this thesis,
presenting the collected data. On the contrary, both their the Synthetic Sentiment Indicators were developed rather by
presentation and interpretation merits ongoing experiments. accident than as a result of purposeful studies, the tests have
One possibility involves using the difference between the two proved that their usefulness in determining market sentiment
lines of the Long vs. Short Sentiment Indicator to create an and forecasting (or rather estimating) the probability of
oscillator that works better in the consolidation period while market extremes emerging is very high. It also may be that the
still highlighting the points of the market being strongly advantage of the described method involves the fact that it is
overbought or oversold. the result of observing a specific market phenomenon instead
of an attempt to adapt a theory to historical data. Nevertheless,
irrespective of the foundations of the described indicators, they
constitute a new and valuable tool in the technical analysis
18 arsenal that permits a better determination of the current
market situation than the broad market indicators deployed to
date. They also help estimate the17market sentiment in the case
 
 

PAGE 10 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Combined Cycles indicator formula:

of stock exchanges that do not release data on the commitment CycleShort:= Fml(“Trailing Stop Short-term”);
levels of respective traders’ groups, thus constituting an CycleMid:= Fml(“Trailing Stop Mid-term”);
alternative to traders’ opinion surveys. While this thesis only CycleLong:= Fml(“Trailing Stop Long-term”);
presents the results of applying the indicators to several
selected indices, the data are promising enough, and the theory SignalLineUP:= Max(Max(CycleLong, CycleMid), CycleShort);
backing them is consistent enough, to support their application SignalLineDW:= Min(Min(CycleLong, CycleMid), CycleShort);
in other markets that are not covered by this study. CyclesComb:= If(C > SignalLineUP, 1, If(C < SignalLineDW, -1, 0));

References UP:= If(CyclesComb = 1, 1, 0);

Ehlers, John F. MESA and Trading Market Cycles: Forecasting and Trading DW:= If(CyclesComb = -1, -1, 0);
Strategies from the Creator of MESA. Wiley, 2002.
Hurst, J.M. The Profit Magic of Stock Transaction Timing. Prentice-Hall, To compute the target aggregate indicators, it is necessary
1970. to aggregate the result of the above indicators for all individual
equities comprised by the given index. It can be achieved in
APPENDIX three ways:
• By using the Security function for every stock price
Indicator Codes for Metastock individually.
• By using the Cum function with the GlobalVars.dll add-in in
Trailing Stop Short-term/Mid-term/Long-term Explorer.
indicators formula:
• By exporting the values of individual indicators to an external
Length:= 1.5; { Mid-term: 3.0 / Long-term: 4.5 }
spreadsheet and importing the aggregated value back to
Volatility:= Length*ATR(5);
Metastock.
StdevOfVolatility:= 3*(1/Sqrt(Length))*Stdev(Volatility,30);
EmaM:= (H+L+C)/3; The result of the above operation should be used in the target
EmaTR:= Volatility + StdevOfVolatility; aggregate Sentiment Indicators. It is marked by the prefix “Agg.”
Stop:= If( PREV < C, in the formulas presented below.
If(( EmaM - EmaTR ) >= PREV, Short-term/Mid-term/Long-term Sentiment
( EmaM - EmaTR ), PREV), Indicators Formula:
( EmaM - EmaTR )); AdjustmentFactor:= 1 {range: 0-2};
StopLong:= Stop;
Sentiment:= ((Agg.CycleShorttermUP*100) - (Agg.
Stop:= If( PREV > C, CycleShorttermDW*(-100))) * AdjustmentFactor;
If(( EmaM + EmaTR ) <= PREV, { Mid-term: Agg.CycleMidtermUP, Agg.CycleMidtermDW
( EmaM + EmaTR ), PREV), Long-term: Agg.CycleLongtermUP, Agg.CycleLongtermDW }
( EmaM + EmaTR ));
StopShort:= Stop; Sentiment.Gray:= If(Sentiment >= -50 AND Sentiment <= 50,
Sentiment, 0);
If(BarsSince(C < Ref(StopLong,-1)) > BarsSince(C > Sentiment.Blue:= If(Sentiment > 50 , Sentiment, 0);
Ref(StopShort,-1)), StopLong, StopShort); Sentiment.Red:= If(Sentiment < -50, Sentiment, 0);

The following values of the Length parameter (illustrating the Sentiment.Gray;

distance between the Stop Trailing line and the stock price) have Sentiment.Blue;
been adopted in this thesis for the respective cycles: Sentiment.Red;
• Short-term: 1.5 Long vs. Short Sentiment Indicators Formula:
• Mid-term: 3.0 AdjustmentFactor:= 1 {range: 0-2};
• Long-term: 4.5
Long.Blue:= (Agg.CycleCombUP*100) * AdjustmentFactor;
Similarly, the relevant parameters for the Mid-term and Short.Red:= (Agg.CycleCombDW*(-100)) * AdjustmentFactor;
Long-term indicators in the following indicator formulas are
provided in the comment. Long.Blue;
Short.Red;
Short-term/Mid-term/Long-term Cycle indicators
formula: The AdjustmentFactor parameter was used in aggregate
SignalLine:= Fml(“Trailing Stop Short-term”); indicators, as it helps adjust the given indicators to the nature
{ Mid-term: Fml(“Trailing Stop Mid-term”) / Long-term: of the given market. It should also be noted that the factor
Fml(“Trailing Stop Long-term”) } does not impact the manner of computing the indicator itself
Cycle:= If(C > SignalLine, 1, If(C < SignalLine, -1, 0)); or its ultimate form but only permits adjusting the absolute
values adopted by the indicators so that they are better adapted
UP:= If(Cycle = 1, 1, 0); visually to overbought and oversold levels.
DW:= If(Cycle = -1, -1, 0);

IFTA.ORG PAGE 11
 IFTA JOURNAL 2021 EDITION

Mohamed M. Khedr, CFTe, MFTA

The Noise Trend Prime Securities

2 Wadi El Nile St., Mohandeseen, Giza, Egypt
By
 Mohamed M. Khedr, CFTe, MFTA
 +201550205296
[email protected]

Abstract with same name. But now, after achieving impressive results on
With the various technical patterns exceeding their minimum stock prices, they are showing a lot of respect to this model.
targets, what should the optimal strategy be for dealing with The technical reports from the various research fields have
such situations? And how can we reduce the risks, especially achieved remarkable results, and the ability of technical
if the breakouts occurred without throwbacks or pullbacks to analysis to measure momentum and predict changes in
the neckline in the head and shoulders patterns or to the apex price trends and technical analysis has been superior to the
in the triangle patterns. This article searches for a new use of knowledge of financial analysis in terms of its ability to predict
amazing Fibonacci retracements to reduce risks and looks at the financial market crises if it occurs reflective analysis patterns
possibility of anticipating peak and trough zones and identifying with distant analytical targets.
oversold and overbought territories after determining the All these reasons led some individual players (noise players)
target zone, especially with the development of the behavior of to pay attention and read technical reports on a daily basis to
some players in the market recently due to many reasons, most help them make decisions.
importantly the strength of technical analysis. I will explain
the impact of this development on trends and the movement of Social media networks
technical patterns. I will also explain how to deal with it. These networks have contributed significantly to the spread
of technical recommendations that related to the market
Acknowledgements and stocks, especially with the intense competition between
I wish to express my sincere appreciation to my advisor, Dr. brokerage companies after the global financial crisis and the
Rolf Wetzer, for his aspiring guidance and for his patience. I need to attract new clients and portfolios.
greatly benefited from his extensive experience. Also, I can’t In addition, social networks have been used as an advertising
ignore the knowledge I got from the all the big financial experts, and marketing method between companies by publishing
especially Professor Charles K. Patrick and Professor Thomas brief technical recommendations that clarify the general and
Bulkowski. Finally, my deep and sincere gratitude to my family expected trends of stock prices in the coming period; here is
for their continuous and unparalleled love, help and support. where the problem showed up.

Introduction Money and fame

As it is known, the financial market experts (Kirkpatrick II Some brokers and users who aren’t specialized at technical
and Dahlquist 2010) classified the market players into three analysis quoted the idea of the mini recommendation and
sections: informed, uniformed and liquidity player. applied it by creating some pages technical in nature to show
They define the noise player as one who has no knowledge various recommendations for buying and selling, which made
about evaluation norms or various analytical methods whether the followers of these pages to reach thousands because the
it is technical or financial, and also, they are the market leaders following two reasons:
for a random movement around the equilibrium price. 1. The ease of speaking as a technical analyst and showing
recommendations—all you have to do is to write the stock
How the Noise Player’s Behavior Has Evolved and name, recommendation, and target (just number) without
the Reasons Behind This attaching a graph or an explanation about when these
To clarify the above questions I will discuss these following positions should be closed (the stop-loss level).
six points: 2. The success of this mini recommendation comes with the
uptrend cycle, which then raises most stocks with different
The power of technical analysis rates (outperforming and underperforming stocks).
The common issue now is that this model is a head and
shoulders model; however, only 50% of its composition has been And with the high salaries of the technical analysts who
completed. Therefore, this model has become the most popular work at brokerage companies or funds, a lot of individuals are
among individuals. And, the interest in this pattern came after motivated to attend basic technical analysis courses, especially
individuals ignored it, which led to great losses, to the extent that after one of the biggest securities companies at Egyptian
some mocked the name of the model because they didn’t know Market hired an administrator of the pages that contain
the reason behind it. Besides, there is a shampoo on the market thousands of members at the company's marketing department.

PAGE 12 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

The technical analyst is not a machine or a system (study and Example one—Discount rates vs. stock market in Argentina
implementation)
Despite the simplicity of technical analysis in terms of study Figure 1. Discount rate in Argentina (Source: Trading
economic | Central Bank of Argentina)
and the ease of passing and obtaining the diploma of technical
analysis after a short period of focus, as well as issuing reports
that contain your technical view of the market, the actual
implementation and market trading is never easy because of
these two reasons:
1. Disadvantage of discretionary system (overtrading,
premature actions, no actions, consistent decision-making
and emotions: the traders often lose money due to their
emotional decisions).
2. Changing positions from buying to selling or vice versa, and
the ability of the technical analyst to discipline, especially if
that process will achieve a real break in his personal account Figure 1 shows that interest rates rose from 20 points to 40
(personal portfolio). points at a rise rate of about 100% in the first half of 2016 and
And here, I would like to refer to the definition of discipline as then tried to retest the 20 point level in 2017 and then enter into
mentioned in the Cambridge Dictionary as “training that makes a sideway range between 20 to 40 points until July 2018. In the
people more willing to have energy and the ability to control first half of August 2018, the interest penetrated the 40-point
themselves.” barrier to reach 60 points in the last half of August 2018, with
So the technical analyst in his early years needs to practice an increase rate of about 50%, and then the sharp rise wave
and fully realize that if the rules are not followed, it will become continued to reach the level of 83 points, which was the peak
more complicated. formed in the second half of 2018. Then, a correction wave will
Therefore, the technical analyst can classify his behavior in start to test the 40 points in the first half of 2019.
his first years as a noise player. Based on this, Argentina’s unprecedented rise in interest
rates from the bottom, which is created around the 20-point
Trend is not always your friend level in December 2015 to 83 points in the last half of 2018,
In order for the technical analyst to decide on the specific which represents a 300% rise in interest rates over the three
strategy for buying or selling, the trend must be determined, years.
and based on this, the strategy is determined too. And so, the The question arises: What is the feedback of Argentina’s
best strategy for the uptrend is to buy and hold, while the financial market to the unprecedented rise in interest rates?
downtrend strategy is to sell the rebound.
In some emerging markets, such as Saudi Arabia, which does Figure 2. The monthly chart of S&P Merval (.MERV) –
Index of Argentina
not have a tool or a mechanism of short selling, and despite the
recent implementation for this tool in some Arab markets, such
as the Egyptian market in 2019 and the Dubai Financial Market,
Abu Dhabi and Qatar market had this license in 2017, that
mechanism was not actually applied by investors for Islamic
religious reasons, which prohibits this type of borrowing.
So the trend may not be the real friend, especially during the
downtrend cycle, and then some individuals adopt the buying
strategy only with the steep down waves trying to make a profit
in the downtrend (countertrend strategy1).

Each rule has an exception (interest rate vs. stock market)

With banks raising interest rates on deposits and investment
certificates, investors are heading to these categories as a safe
asset, away from stocks, which are a risky investment. Raising
these rates will raise the cost to borrowing companies, and some
of it may be moving to accelerate their expansion plans and new
projects. Figure 2 shows an example of the ascending triangle that was
In the end, this will reduce the stock price that listed at the penetrated in the first half of 2016, and after that, the index
market and in turn will lead to a decline in the market and its direction changed from the sideway trend to the uptrend, beside
indices, so we have two technical rules: the first one is three its rising from the support level of 9,200 points, at which the
steps and a stumble,2 and the second is two tumbles and a jump.3 bottom formed in January 2016, to 35,500 points at the peak,
But each rule has an exception. I will show two examples of which formed in January 2018, with a rise rate about 285% in
the positive relationship between interest rates and the stock two years.
market.

IFTA.ORG PAGE 13
 IFTA JOURNAL 2021 EDITION

How did that happen? The answer is the flotation (the approximately the nominal interest rate minus the inflation
exception) and the need to determine the real interest rate. rate. The real interest rate measures the compensation for
expected losses due to default and regulatory changes as well
Figure 3. The monthly chart of USD/ARS – U.S. dollar/ as measuring the time value of money; they differ from nominal
Argentinean peso
rates of interest by excluding the inflation compensation
component.
For example, if an investor was able to lock in a 5% interest
rate for the coming year and anticipated a 2% rise in prices, they
would expect to earn a real interest rate of 3%.
How to measure:
Real interest rate = Nominal interest rate – expected inflation
rate.
Example two—Discount rates vs. stock market in Egypt
Figure 5. Discount rate in Egypt (source: investing.com)

Figure 3 shows the rise of the dollar's value against the

Argentine peso from 9.80, which is the bottom created in
December 2015, to the peak level of 41.70 formed in the second
half of 2018, with a rise rate of 325%.
I will clarify the effect of this sharp decline in the currency of Figure 5 shows the rise in interest rates from the level of 9.75
Argentina (peso) against the dollar on inflation rate. at the end of December 2015 to reach the level of 19.25, which is
the peak formed in July 2017, with an increase rate of about 97%.
Figure 4. The bar chart of the inflation rate in Argentina And in the beginning of 2018, specifically in November, interest
(Source: tradingeconomics.com, Instituto Nacional de
rates started in a correction wave to reach the level of 12.75 by
Estadística y Censos (INDEC))
the end of 2019.
Figure 6. EGX 30 index (.EGX30) – Egyptian Stock Index

Figure 4 shows that the high rates of inflation in Argentina

from the level around 15% at the beginning of 2016, reaching
around 50% by the end of 2018. This means that inflation
rates increased in that period by about 233% as a result of the
currency flotation decision, which in turn led to interest rates
increasing and controlling inflation.
And this is why we can say that both three steps and a
stumble and two tumbles and jump rules have exceptions, and
the exception here is the sharp change in the currency value and Figure 6 shows the rise of the Egyptian market index from
the need to determine the real interest rate. the level of 5,550, which is the bottom formed in January 2016,
Therefore I would like to highlight more about the real reaching to the level of 15,721, which is the highest level of
interest rate, which is the rate of interest an investor, saver the index in November 2018, with a rise rate about 183%. And
or lender receives (or expects to receive) after allowing despite the gradual decline of interest rates, this decline came
for inflation. It can be described more formally by the with a violent correction wave in the index, reaching the support
Fisher equation, which states that the real interest rate is level of 12,000 points by the end of 2018.

PAGE 14 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Figure 7. The monthly chart of USD/EGP – (US dollar/

Egyptian pound) Materials and Methods
This effect appears through two points: the noise trend and
changes in the psychology of some technical patterns.

The Noise Trend

Definition
The noise trend is defined as resulting from a random price
movement and sharp fluctuations from a series of violent
uptrend waves in both the short and medium term, followed
by waves of sharp downtrends, which will erase the long-term
features of the main trend (more than a year). Beside this, that
trend is similar in its shape and composition to the sideway
trend but on a larger scale.
This trend caused sharp fluctuations in stock prices among
traders, which in turn contributed to the appearance of a new
kind of market trend.
Figure 7 shows the dollar rise against the Egyptian pound In general, the oscillating trend aimed at a price that moves
from the level of 9, which is the bottom formed in November between support zones and other resistance and not at specific
2016 (floatation decision), to the level of 19.19, which is the peak levels.
formed by the end of 2016, with a rise rate of the dollar against
the pound, which reached about 121% in two months, to appear Profiting from the noise trends: Hot money stole money
in 2017 as a correction wave that pushed the dollar down around In the noise trend, the hot money makes a lot of money
the level of 15.69, which is the bottom formed in February 2017. through large-scale price movements that may reach its upward
wave rates exceeding 100% during a period not exceeding two
Figure 8. The bar chart of the inflation rate in Egypt years after a sharp drop of waves that may range between 60
(Source: tradingeconomics.com | Central Bank of Egypt)
and 80%. I will show some historical statistics of this and the
results from these examples. (See Appendix A)
Figure 9. Egyptian Transport & Commercial Services Co.
SAE (ETRS.CA) – Equity in Egypt

Figure 8 shows the continuation of inflation rates in Egypt

from the level of 13.56 in October 2016, reaching the level
of 31.916, which is the peak formed in August 2017, with an
increase reached to about 135%.

Figure 9 shows how the price has been moved between the
wide range levels (2.50:16.00) during the consecutive violent
waves.

IFTA.ORG PAGE 15
 IFTA JOURNAL 2021 EDITION

Table 1. Historical statistics of the noise trend for (ETRS.CA)

Date
The wave number Price range From To Percentage (%)
The first violent wave 3.47:16.27 July 2007 September 2008 369%
The second violent wave 16.27:4.63 September 2008 October 2008 71.50%
The third violent wave 4.63:16.00 October 2008 October 2009 245.50%
The fourth violent wave 16.00:2.43 October 2009 June 2012 85%
The fifth violent wave 2.43:8.51 June 2012 September 2012 250%
The sixth violent wave 8.51:3.41 September 2012 June 2013 60%
The seventh violent wave 3.41:9.74 June 2013 October 2014 185%
The eighth violent wave 9.74:2.80 October 2014 January 2016 71%
The tenth violent wave 2.80:15.96 January 2016 January 2018 470%

Note that these violent waves were measured from the troughs (the low price) to the peaks (the high price).

Statistical results:
Frequent historical price levels produced violent upward waves (3.50-4.60-2.40-3.40-2.80), so we will consider the levels between
2.40 and 4.60 as a significant zone. On the other hand, there were historical price levels that produced sharp declining waves (16.00-
8.50-10.00), so we will consider the levels between 8.50 and 16 as a significant zone.
Figure 10. Ezz Steel (ESRS.CA) – Equity in Egypt Figure 10 shows how the price has been moved between the
wide sideways range (6.00:46.00) levels during the consecutive
violent waves.

Table 2. Historical statistics of the noise trend for (ESRS.CA)

Date
The wave number Price range From To Percentage (%)
The first violent wave 7.42:46.36 May 2005 February 2006 525%
The second violent wave 46.36:14.64 February 2006 June 2006 68%
The third violent wave 14.64:38.45 June 2006 April 2008 162.50%
The fourth violent wave 38.45:5.84 April 2008 February 2009 85%
The fifth violent wave 5.84:26.00 February 2009 April 2010 345%
The sixth violent wave 26.00:3.28 April 2010 December 2011 87%
The seventh violent wave 3.28:19.50 December 2011 September 2014 494.50%
The eighth violent wave 19.50:6.11 September 2014 January 2016 68.50%
The tenth violent wave 6.11:30.79 January 2016 April 2019 404%

Statistical results:
Frequent historical levels produced violent upward waves (7.50-6.00-3.00), so we will consider the levels between 3.00 and 7.00 as a
significant zone. On the other hand, there were historical levels that produced sharp declining waves (46.00-38.50-19.50-31.00-26), so
we will consider the levels between 19.50 and 46.00 as a significant zone.

PAGE 16 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Figure 11. Misr Chemical Industries (MICH.CA) – Equity Figure 11 shows how the price has been moved between the
in Egypt wide range levels (3.00:16.00) during the consecutive violent
waves.

Table 3. Historical statistics of the noise trend for MICH.CA

Date
The wave number Price range From To Percentage (%)
The first wave 3.17:16.58 December 2003 January 2006 423%
The second wave 16.58:4.46 January 2006 July 2006 73%
The third wave 4.46:14.41 July 2006 May 2008 223%
The fourth wave 14.41:2.77 May 2008 October 2008 80.77%
The fifth wave 2.77:9.31 October 2008 September 2009 236%
The sixth wave 9.31:3.00 September 2009 November 2011 67.77%
The seventh wave 3.00:7.88 November 2011 June 2014 162.66%
The eighth wave 7.88:2.80 June 2014 November 2016 64%
The tenth wave 2.80:15.73 November 2016 January 2018 462%
The eleventh wave 7.88:2.81 June 2014 November 2016 64%
The twelfth wave 2.81:15.73 November 2016 January 2018 460%

Statistical results:
Frequent historical price levels produced violent upward waves (3.00-4.50), which will be considered as a significant zone. On the
other hand, there were historical price levels that produced sharp declining waves (16.00-14.50-8.00-9.00), so we will consider the
levels between 14.00 and 16.50 as a significant zone.

More patterns, more noise Figure 12. El Nasr Agricultural Crops (ELNA.CA) – Equity
The appearance of the patterns with targets higher than the in Egypt
current price intensively increases the chances of having this
type of trend.
Figure 12 shows another example of the noise trend between
levels of 5.50 and 42.00. Note that inside this trend there were
many technical patterns, such as a triple top pattern with
maximum target at the 5.50 level and counter target at the
30.50 level. After that, the stock witnessed a double bottom
pattern with a maximum target at the 42.00 level and the
counter target at the 17.00 level. The result is that the presence
of many technical patterns can cause this type of trend.

IFTA.ORG PAGE 17
 IFTA JOURNAL 2021 EDITION

How to reduce the sound of noise The Technical Patterns

The answer is to identify the zones of support and resistance This behavior had an impact on the different technical
based on the historical price movements and then determine the patterns, especially the reversed trend, and its effect is divided
risks of trading in the direction of the price, as the risk of selling into two parts:
increases near the historical support areas, even if the price 1. Re-testing: It became common not to re-test the neck line in
trend is the downtrend until this zone is violated, while the risks the most popular model (head and shoulders) besides the lack
of buying increases near the historical resistance areas even if of return to the apex in most triangle models.
the price trend is the upward trend until this zone is penetrated. 2. The Accurate Determination of Target: The lack of a careful
More confidence in the technical indicators signals related to targeting, and here I mean the maximum target, so the price
divergences or momentum, especially that which is given near can easily exceed the minimum target and at other times
these areas of historical support or resistance. There is more the maximum target. Besides the model target can be found
reservation when the validity of penetration is tested, such somewhere between the minimum and the maximum target.
as the penetration rules (3: 5%), where it is preferable to use And now, I will show some important statistics of a financial
a penetration rule of 5% in this type of trend with sharp price markets expert (Bulkowski 2005), related to the most important
fluctuations to avoid the false breakouts. technical patterns presented in his book Encyclopedia of Chart
Patterns.
Table 4. General statistics in a bull market (Source: Encyclopedia of Chart Patterns, 2nd Edition)
Number of Pullback or Average Rise or Rise or Decline Over Change After Trend
Formations Formations Throwback Decline (%) 45.00 (%) End
Double bottoms, Adam 206 64.00% 35.00% 58.00 or 28.00% -33.00%
& Adam
Double tops, Adam & 188 61.00% -19.00% 9.00 or 5.00% 54.00%
Adam
Head and shoulders 640 50.00% -22.00% 34.00 or 5.00% 51.00%
tops
Head and shoulders 554 45.00% 38.00% 187.00 or 34.00% -31.00%
bottoms
Rectangle bottom 115 53.00% 46.00% 48.00 or 42.00% -28.00%
Descending triangles 312 37.00% 47.00% 125.00 or 40.00% -30.00%
Ascending triangle 663 57.00% 35.00% 200.00 or 24.00% -29.00%
Symmetrical triangle 476 54.00% 31.00% 144.00 or 30.00% -31.00%
Pennants 173 47.00% 25.00% 27.00 or 16.00% -25.00%
Failing wedge 245 56.00% 32.00% 80.00 or 33.00% -28.00%

Glossary explanation:
Change after trend end (See Appendix B): In the double 2. Amazing results from the component models, such as
bottom, Adams & Adam, once prices top out at the ultimate high, triangles, rectangles, and head and shoulders bottom,
then what happens? They tumble about 33%. In a bear market, for which targets exceed 45%, and that is a high rate and
the decline gives up all the gains and more. For aggressive indicates the ease of exceeding the minimum target with the
traders in a bear market, wait for prices to peak and then short possibility of exceeding the maximum of the pattern target,
the stock and ride it down. or that target is somewhere between the minimum and the
In an analysis of this previous statistical data we found that: maximum of a perfect model target.
1. About 50% or more of these models (head and shoulders 3. When comparing the average rise or decline ratios with
top, head and shoulders bottom, the descending triangle change after trend end ratios, we will find the following:
and pennant) didn’t rebound to the neckline or to the apex
after penetration, which refers to the strategy of waiting
for the throwback or pullback to buy or to add, and to short
doesn’t go with many technical patterns, which leads to the
need for another strategy to avoid the possible unexpected
throwbacks or pullbacks.

PAGE 18 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Table 5. Comparing the statistical results (average rise or decline/change after trend end)
Return/Risk Ratio of the Trading
in the Direction of Pattern’s
Formations Average Rise or Decline (?) Change After Trend End (%) Breakout
Double bottoms, Adam & Adam 35.00% -33.00% 1:1
Double tops, Adam & Adam -19.00% 54.00% 1:2.8
Head and shoulders tops -22.00% 51.00% 1:2.3
Head and shoulders bottoms 38.00% -31.00% 1:0.80
Rectangle bottom 46.00% -28.00% 1.65:1
Descending triangles 47.00% -30.00% 1.56:1
Ascending triangles 35.00% -29.00% 1.20:1
Symmetrical triangles 31.00% -31.00% 1:1
Pennants 25.00% -25.00% 1:1
Failing wedge 32.00% -28.00% 1.14:1

A. Some models, such as double bottom, head and shoulders Target Zone
bottom, symmetrical triangle, and pennant, with average The answer is to determine a target zone for each pattern
of rise or decline target ratios are equal to the change where the risks of trading in the direction of the pattern's
after trend end ratios. This means that after achieving and breakout are higher than the reward.
reaching the target, the stock can witness a strong rebound
that may greatly reduce the profits achieved from the target’s Definition
return. As breakouts from the support or resistance run to the next
B. Another form of models like the double top and head and zone of support or resistance, the breakout of minimum price
shoulders top which their change after trend end ratios target in the chart patterns run to the maximum price target,
exceeded more than the double of the average rise or decline which represent 100% of the previous trend wave. This gives
ratios. This means that after achieving and reaching the the investors a price objective from trading below or above
target, the stock can witness a strong rebound that may erase the minimum target while the risk is the amount that the price
all the profits achieved from the target’s return. may go against the pattern breakout to bounce or to correct
And this indicates to the following: (countertrend bounce) and then risk/reward ratio can be
1. Trend is not always your friend (sometimes the trading calculated?
risks increase in the direction of the current price).
2. These models had three price targets: the minimum of How to Measure?
target (average rise or decline), the maximum of target Calculating the risk/reward ratio of trading below or above
(rise or decline over 45%) and the counter target (change the minimum target using Fibonacci retracement:
after trend end). Step number one: The return ratio below or above the
3. The need and importance of determining where the minimum target is the difference between the entry price and
stock will peak or trough after the pattern breakouts to the maximum of target divided by the entry level *100.
maximize profits from these patterns by benefit of the
counter target.
4. The importance of managing the risks of trading in the
direction of the pattern breakout, especially when the
price exceeds the minimum target. Step number two: The percent (38.20%) of Fibonacci
How can we reduce the risks of the counter-pattern targets retracements is used to identify a countertrend bounce from the
using Fibonacci retracement? When is stopping trading in the last peak or trough within the pattern to the maximum target of
direction of the breakout a must for the conservative investors? that pattern.
Step number three: The risk ratio of trading in the direction
of the patterns’ breakout (counter target) is the difference
between the minimum correction (countertrend bounce) and
the entry level*100.

Step number four: Comparing the risk ratio with reward ratio
to determine the target zone where the risk of correction is
higher than the reward of reaching the maximum target.

IFTA.ORG PAGE 19
 IFTA JOURNAL 2021 EDITION

Table 7. How to calculate the risk ratios of trading below

Implementation the minimum target

Figure 13. UAC of Nigeria PLC (UACN.LG) – Equity in Minimum Rebound

Nigeria to the 38.20% of
Fibonacci, Which Is
Entry Level (Short) Supposed to Start
Below the Minimum From Maximum
Target Target Risk Ratio (%)
33.00 40.80 23.63%
30.00 40.80 36%
29.00 40.80 40.68%

Based on step number four: risk/reward ratio comparison.

Table 8. Risk/reward ratio comparison of trading below
the minimum target
Entry Level (Short)
Below the Minimum
Target Risk Ratio (%) Reward Ratio (%)
33.00 23.63% 27%
30.00 36% 20%
Figure 13 shows that the stock has witnessed a head and 29.00 40.68% 17%
shoulders pattern; the minimum target was the level of 35.00
and the maximum target was the 24.00 level. It also shows how The results in Tables 7 and 8 show the increase in the risk of
the price easily exceeded the minimum target at the 35.00 level trading in the direction of a pattern’s breakout compared to the
to hit the 26.00 level, and then there was a strong bounce wave expected return of reaching to the maximum target starting
from it by 69% to the 44.49 level (counter target). from the level of 30.00, so our target zone that helps us to
reduce the risks and avoid the counter target will be between
The processes to handle with this situation: the 30.00:24.00 levels, where the stop selling is a must for the
Based on step number one: conservative investors based on the risk ratios exceeding the
Calculating the reward ratios of reaching the maximum reward ratio.
target at the 24.00 level:
Figure 14. Anel Elektrik Proje Taahhut ve Ticaret AS
Table 6. How to calculate the return ratios of reaching (ANELE.IS) – Equity in Turkish
the maximum target
Entry Level (Short)
Below the Minimum
Target Maximum Target Return Ratio (%)
33.00 24.00 27%
30.00 24.00 20%
29.00 24.00 17%

Based on step number two:

The countertrend bounce is near the level of 40.80, which
represents 38.20% of Fibonacci retracement which was taken
from the distance between the last pattern’s peak at the level of
68.00 and the maximum target at the level of 24.00.
Based on step number three:
Calculating the risk ratios of the bounce wave (38.20% of
Fibonacci retracement), which is supposed to start from the
maximum target 24.00: Figure 14 shows the double top reversal pattern that was
violated in March 2018, and it had a minimum target at 2.50
and a maximum target at 1.15. The graph also shows how the
minimum target was easily broken until it succeeded in creating
a bottom at 1.50 to start a quick initial bounce reached to the
level of 2.05.

PAGE 20 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

The processes to handle with this situation: The processes to handle with this situation:
Based on step number 4: Based on step number one: Calculating the reward ratio of
reaching the maximum target at the 147.50 level:
Table 9. Risk/reward ratio comparison of trading below
the minimum target Table 10. How to calculate the return ratios of reaching
the maximum target
Entry Level (Short)
Below the Minimum Entry Level (Long)
Target Risk Ratio (%) Reward Ratio (%) Abovethe Minimum
2.00 9% 42.50% Target Maximum Target Return Ratio (%)
1.80 21% 36% 90 147.50 63.88%
1.60 36% 28% 100 147.50 47.50%
1.50 45% 23% 120 147.50 22.92%
130 147.50 13.46%
The results in Table 9 show the increase in the risk of trading
in the direction of a pattern’s breakout compared to the
expected return of reaching to the maximum target starting Based on step number two: The countertrend bounce will be
from the level of 1.60, so our target zone that helps us to reduce near the 106.00 level, which represents 38.20% of Fibonacci
the risks and to avoid the counter target will be between retracements (this correction ratio was measured from the
the 1.60:1.15 levels, where the stop selling is a must for the right shoulder at the level of 39.00 to the maximum target at the
conservative investors based on the risk ratios being higher level of 147.50).
than the reward ratios. Based on step number three: Calculating the risk ratios of
the bounce wave (38.20% of Fibonacci retracement), which is
Figure 15. Brent Oil Futures (LCOc1) – Commodities supposed to start from the maximum target 147.50.
Table 11. How to calculate the risk ratios of trading above
the minimum target
Minimum Rebound
to the 38.20% of
Fibonacci, Which Is
Entry Level (Long) Supposed to Start
Above the Minimum From Maximum
Target Target Risk Ratio (%)
90 106 -
100 106 -
120 106 11.66%
130 106 18.46%

Based on step number four: Risk/reward ratio comparison:

Figure 15 shows a head and shoulders bottom has been Table 12. Risk/reward comparison of trading above the
minimum target
formed since March 2009 after penetrating the 50.00 barrier,
and its targets were the minimum target around 68.00 and Entry Level (Long)
the maximum target (100% of the previous downside wave) Above the Minimum
at 147.50. The graph also shows how the price easily exceeded Target Risk Ratio (%) Reward Ratio (%)
the minimum target to hit the level of 128.50, which is the 90 - 63.88%
peak formed in the month of March 2013. Then, this rally was 100 - 47.50%
followed by waves of decline until it reaches a level of 88.50, 120 11.66% 22.92%
with a decrease by 30% from the ultimate high, which we called
130 18.46% 13.46%
the counter target.

Table 12 shows the increase in the risk of trading in the

direction of a pattern’s breakout compared to the expected
return of reaching the maximum target starting from the
level of 130.00, so our target zone that helps us to reduce
the risks and to avoid the counter target will be between the
130.00:147.50 levels, where the stop buying is a must for the
conservative investors based on the risk ratios exceeding the
reward ratios.

IFTA.ORG PAGE 21
 IFTA JOURNAL 2021 EDITION

Figure 16. Crude oil (CLc1) – Commodities Figure 17. Copper future (HGc1) – Commodities

Figure 16 shows a symmetrical triangle has been formed since Figure 17 shows the double top reversal pattern that has been
October 2014 after violating the support level of 85.00, and its violated in September 2011, and it had a minimum target at 3.10
targets were the minimum target around the 41.00 level and the and a maximum target at level of 1.25. The graph also shows
maximum target (100% of the previous upward wave) was at the how the minimum target was easily broken until it succeeded
32.50 level. The graph also shows how the price easily exceeded in creating a bottom at 1.95 (ultimate low) to start a bounce
both the minimum and maximum targets to hit the level of reached to the level of 3.30 (counter target).
26.00, which is the trough formed in the month of January 2016.
Then, this decline was followed by rising waves until it reached The processes to handle with this situation:
a level of 75.50, with an increase by 190% from the ultimate low, Based on step number two: the countertrend bounce will
which we called the counter target. be around the level of 2.50, which represents the 38.20% of
Fibonacci retracements (this correction ratio was measured
The processes to handle with this situation: from the second peak of the head and shoulders pattern at the
Based on step number two: The countertrend bounce will be level of 4.50 to the maximum target at the level of 1.25).
near the level of 61.00, which represents 38.20% of Fibonacci Based on step number four:
retracement (this correction ratio was measured from the last
peak of the symmetrical pattern at the level of 108.00 to the Table 14. Risk/reward ratio comparison of trading below
the minimum target
maximum target at the level of 32.50).
Based on step number four: Entry Level (Short)
Below the Minimum
Table 13. Risk/reward comparison of trading below the Target Risk Ratio (%) Reward Ratio (%)
minimum target 2.50 - 50%
Entry Level (Short) 2.30 8.69% 45.65%
Below the Minimum 2.00 25% 37.50%
Target Risk Ratio (%) Reward Ratio (%)
1.90 31.58% 34.21%
40.00 52.50% 18.75%
1.80 38.88% 30.55%
35.00 74% 7%

The results above show the increase in the risk of trading in Table 14 shows the increase in the risk of trading in the
the direction of a pattern’s breakout compared to the expected direction of a pattern’s breakout compared to the expected
return of reaching the maximum target starting from the level return of reaching the maximum target starting from the level
of 41.00, so our target zone that helps us to reduce the risks of 1.90, so our target zone which helps us to reduce the risks and
and to avoid the counter target will be between the 41.00:32.00 to avoid the counter target will be between the 1.90:1.25 levels,
levels, where the stop selling is a must for the conservative where the stop selling is a must for the conservative investors
investors based on risk ratios being double the reward ratio. based on the risk ratios being equal to or increasing more than
the reward ratio.

PAGE 22 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Puzzle 3. Ternium Argentina SA (TXAR.BA) – Equity in

Focus on Failure Argentina

When Does This Method Fail to Achieve Its Main

Purpose of Reducing the Risk of Trading in the
Direction of the Pattern’s Breakouts?
From the research, we found that using this method to reduce
the risk of the counter target is consistent with most cases in
which the patterns did not fail because it is based on the rules
and foundations of technical analysis, and they are correction
ratios (Fibonacci retracement) and the risk/reward ratio, but
what if the maximum target is unknown, and this is happening
when the pattern’s targets reaches a new historical level (all-
time high)?
Puzzle 1. McDonald’s (MCD) – Stock in USA

Puzzle 3 shows that the broadening formation has been

formed after breaking the barrier of 17.40 with minimum target
at the 24.00 level, but the maximum target is unknown (the
stock moved at new historical levels).

How to Overcome This Problem to

(Source: investing.com)
Define Our Target Zone
The solution for these puzzles is to use the amazing Fibonacci
Puzzle 1 shows a bullish triangle pattern has been formed retracements, especially the 423.00% that will be taken into
after penetrating the barrier at 102.00 levels with its minimum consideration as a significant resistance level (maximum
target at approximately 122.00 levels. The maximum target is target), and it is calculated by taking the distance from the last
unknown (the stock was moving in new historical levels). peak to the last bottom within the pattern.
Puzzle 2. Sociedad Comercial Del Plata (COME.BA) – The Implementation
Equity in Argentina
Puzzle 4. Solving the puzzle number 1 McDonald’s (MCD)
– Equity in USA

(Source: investing.com)

Puzzle 4 shows how the maximum target of the pattern is

determined using the ratio 423% of Fibonacci retracement,
especially if the pattern is at new historical levels, and it is
calculated by taking the distance from the last peak at the
level of 101.40 to the last bottom at the level of 87.50 within the
Puzzle 2 shows an ascending triangle has been formed after pattern, which is in this case at the 145.00 level, which will be
penetrating the resistance level of 1.95 with the minimum target considered as a maximum target. Then we can find our target
at the 2.65 level, but the maximum target is unknown (the stock zone.
moved at new historical levels).

IFTA.ORG PAGE 23
 IFTA JOURNAL 2021 EDITION

The process to find our target zone: The process to find our target zone:
Based on step number four: Based on step number four:
Table 15. Risk/reward ratio comparison of trading above Table 16. Risk/reward ratio comparison of trading above
the minimum target the minimum target
Risk Ratio (%) of the Reward Ratio (%) of Risk Ratio (%) of the Reward Ratio (%)
Expected Counter Targeting the 145.00 Expected Counter of Targeting the
Entry Level (Long) Target at 123.00 Level (423.00% Entry Level (Long) Target at 2.65 3.37 Level (423.00%
Above the Minimum (38.20% of Fibonacci of Fibonacci Above the Minimum (38.20% of Fibonacci of Fibonacci
Target Retracement) Retracement) Target Retracement) Retracement)
125.00 1.60% 16.00% 2.70 1.85% 24.80%
130.00 5.38% 11.50% 2.80 5.35% 20.35%
132.00 6.82% 9.85% 3.05 13% 10.50%
133.00 7.50% 9.00% 3.10 14.50% 8.71%
134.00 8.21% 8.20%
Table 16 shows that our target zone will be between the
Table 15 shows that our target zone will be between the 3.05:3.37 levels, where the reward ratios are less than the risk
134.00:144.00 levels, where the reward ratios are equal to or less ratio, and the counter target or strong correction is expected to
than the risk ratio, and the counter target or strong correction is happen near this zone.
expected to happen near this zone.
Puzzle 6. Solving the puzzle number 3 Ternium
Puzzle 5. Solving the puzzle number 2 Sociedad Argentina SA (TXAR.BA) – Equity in Argentina
Comercial del Plata (COME.BA) – Equity in Argentina

Puzzle 6 shows how the maximum target of the pattern is

determined by using the ratio 423% of Fibonacci retracement,
Puzzle 5 shows how the maximum target of the pattern is and it is calculated by taking the distance from the last peak at
determined using the ratio 423% of Fibonacci retracement, and it the level of 15.20 to the last bottom at the level of 10.45 within
is calculated by taking the distance from the last peak at the level the pattern, which is in this case at the 30.60 level, which will
of 2.77 to the last bottom at the level of 1.67 within the pattern, be considered as a maximum target. Then we can find our
which is in this case at the 3.37 level, which will be considered as target zone.
a maximum target. Then we can find our target zone. Based on step number four:
Table 17. Risk/reward ratio comparison of trading above
the minimum target
Risk Ratio (%) of the Reward Ratio (%) of
Expected Counter Targeting the 30.60
Entry Level (Long) Target at 22.90 Level (423.00%
Above the Minimum (38.20% of Fibonacci of Fibonacci
Target Retracement) Retracement)
26.00 11.92% 17.69%
27.00 15.19% 13.33%
28.00 18.21% 9.82%

PAGE 24 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Table 17 shows that our target zone will be between the References
27.00:30.70 levels, where the reward ratios are less than the risk Bulkowski, Thomas N., Encyclopedia of Chart Patterns. New Jersey:
ratio, and the counter target or strong correction is expected to Wiley 2 edition, 2005.
happen near this zone. Kirkpatrick II, Charles D. and , Julie R. Dahlquist, Technical Analysis:
Stock prices can go up endlessly. What will we do if the The Complete Resource for Financial Market Technicians. New
supposed maximum target (423% of Fibonacci) is penetrated? Jersey: FT Press, 2010.
After breaking the 423%, which is considered as a strong
resistance level, we will use the penetration rule of 5% for Notes
confirmation and then the stop loss has to activate (re-entry). 1 A countertrend strategy is a trading method that attempts to make
small gains by trading against the current trend. Traders also refer
Results to the practice as countertrend trading.
2 Three steps and a stumble, which states that “whenever the
Determining the risk/reward ratio enables us to manage the
Federal Reserve raises either the federal fund target rate, margin,
risks and in case of technical patterns that exceed the minimum
requirement or reserve requirements three consecutive times
target, the reward is the maximum target and the potential risks without a decline, the stock market is likely to suffer substantially.”
are the lowest rebound wave, which we suppose is a correction 3 Two tumbles and jump, which is based on looking for two
rate of 38.20% of the Fibonacci ratios, which we also assume will consecutive declines, or tumbles, in the federal fund target rate and
begin after reaching the maximum target. Ultimately, price levels reserve requirements, and then the stock market rises. But every
where the risk is higher than the reward rate are called target rule has an exception.
zone make technical patterns more profitable for investors by
taking advantage of the countertrend. Software and Data
All Data obtained from MetaStock Version 17.1.1932602,
Discussion and I would like to thank each of the sites (Investing.com &
As for the two rules (two tumbles and jump and three steps Tradingeconomics.com).
and a stumble), we should clarify how effective and important
these rules are in developed markets. For emerging markets, Appendices
where currency rates are changing dramatically and where
inflation rates are affecting them, there is an urgent need to Appendix A
determine the real interest rate and the real inflation rate
as well. There are some countries that do not give real data Figure of Access Bank (ACCESS.LG) – Equity in Nigeria
on annual inflation rates so as not to stir up public opinion,
especially when the government makes difficult economic
reforms to the people, as they take a long time to show their
results.

Conclusion
The strength of technical analysis and its spread among
dealers had a significant impact on price movement, especially
in emerging markets, which in turn led to the creation of a new
type of trend whose waves are characterized by violence. This
made it difficult to accurately determine targets (maximum
targets) in the more common technical patterns, and this
arranges the difficulty of calculating the return/risk ratio of
the trade in the direction of the pattern breakouts, especially
if the price exceeds the minimum target. Technical analysis did
not leave anything. Fibonacci retracements were the effective
tool to overcome the violent price movements. The 38.20% of The monthly chart above is another example of a noise trend
Fibonacci retracements is used as a trailing stop to determine (similar to a sideways trend), which was moving between the
the target zone, while the ratio of 423% is used to find the levels of 4.00:12.50. To reduce this sound of noise, it is necessary
unknown maximum target. The target zone has become an to identify the support area, where the selling risks increase,
urgent necessity recently because it supports reducing the and the resistance zone, where the buying risks increase, from
risks, which is a major goal of technical analysis, and not only to the historical price movements of the stock.
maximize profits.

IFTA.ORG PAGE 25
 IFTA JOURNAL 2021 EDITION

Appendix B
Table for Explanation (Source: Encyclopedia of Chart Patterns, 2nd Edition)
Formation Explanation
(Throwbacks OR Pullbacks and change after trend end)
Double Bottoms, Adam & Adam *Since a throwback occurs 64% of the time in a bull market, you can initiate a position after the
throwback is complete or add to your position.
*Change after trend ends. Once prices top out at the ultimate high, then what happens?
They tumble about 33%. In a bear market, the decline gives up all the gains and more. For aggressive
traders in a bear market, wait for prices to peak then short the stock and ride it down.
Double Tops, Adam & Adam *Change after trend ends. Once a price reaches the ultimate low, it soars by climbing 54% in a bull. Thus,
if you can determine when the price trend changes from down to up, buy.
Head and shoulders tops *Change after trend ends. After a price reaches the ultimate low, it soars, climbing 51% in a bull market. If
you can determine when the trend changes, then buy the stock and hold on.
Head and shoulders bottoms *Change after trend ends. After a price reached the ultimate high, it tumbled by over 30%. That behavior
is a good reason for selling buy-and-hold positions.
Descending triangles *Change after trend ends. Once a price reaches the ultimate high or low, what happens? For upward
breakouts, prices tumble about 32%, on average. Thus, if you can determine when the price has changed,
even if you are late, you can still profit from it.

PAGE 26 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Momen Atef El Shayal, CFTe, MFTA

Determination of Time Target Zones for SIGMA Capital

37 Gameat El Dowal El Arabeya Street
Price Targets of Classic Price Patterns Mohandessinm, Giza 12411
Egypt
By Momen Atef El Shayal, CFTe, MFTA +201202257088
[email protected]

Abstract Technical Analysis Explained (Pring), and Technical Analysis: The

With the claim that it is possible to begin to scientifically Complete Resource for Financial Market Technicians (Kirkpatrick
answer the difficult question of how long it takes to get this and Dahlquist), tried to simplify them; moreover, others, like
potential profit, this paper came to uncover a new dimension Encyclopedia of Chart Patterns (Bulkowski) tried to collect
of some classic price patterns that reflect information about most of them. Although some of the classic price patterns can
the approximate duration within which their price targets are be categorized as reversal patterns and others can do a dual
expected to be achieved. job as reversal and continuation patterns, in this paper, they
A new method that uses concepts of linear regression, and the were categorized as bullish and bearish ones according to the
standard deviation is introduced here to determine time target breakout direction (Figure 1).
zones for the traditionally determined price targets of classic
price patterns. Figure 1. Bullish and bearish price patterns
Seven classic price patterns were tested to prove the thesis.
They include double top (and bottom); triple top (and bottom);
bullish and bearish: rectangle; ascending triangle; descending
triangle; symmetrical triangle; and head and shoulders.
I believe that if we could get such information, we can make
more effective and well-planned investment decisions and it
will be much easier to choose from various coinciding trading
opportunities.

Introduction
In fact, most, if not all, efficient investment decisions are
taken after answering two main questions; the first is how
much is the expected return? And the second is how much
time is expected to get that return? Since the decision to invest
in financial markets should be an efficient one, it needs to be
supported by information about the expected time that is
required to earn an expected return. At the same time, most
technical analysis theories focus on studying price action to
determine future price targets; however, they don’t focus with
the same degree on determining the time targets of these price
targets.
For instance, classic price patterns are a cornerstone in the
body of technical analysis; they reflect how history repeats
itself throughout frequent battles—based on fear and greed—
between buyers and sellers. These battles appear on charts as
recognizable repetitive sideways price patterns, and an easy
minimum price target can be determined once a price breakout
occurs by measuring the largest vertical distance between the On one side, theoretically, the last price swing in the price
pattern’s boundaries and projecting it from the breakout point. pattern represents the beginning of imposing control from the
So far it is good, but what about the time it takes to achieve this warrior who finally wins the trading battle. So, the focus on the
price target?!! last price swing can certainly reveal something worthy about
Actually, there are a lot of classic price patterns with different the remaining power of the victor.
degrees of fame. Some books, such as Technical Analysis of On the other side, the flashing sentence that Perry J. Kaufman
Stock Trends (Edwards, Magee and Bassetti) and Trading Classic included in the 5th edition of his book Trading Systems and
Chart Patterns (Bulkowski), tried to explain them in details; Methods—“A time series is not just a series of numbers, but
others, like Technical Analysis of Financial Markets (Murphy), ordered pairs of price and time.”—made me think about the fact

IFTA.ORG PAGE 27
 IFTA JOURNAL 2021 EDITION

that we already have two axes (price and time), and completed Materials and Methods
classic price patterns enable us to determine a future point on
the price axis, so there must be a way to get the opposite point Data and Period of Investigation
on the other axis (the time axis). To test the thesis presented in this paper, daily historical data
After gathering all together with the aid of a simple of three different liquid financial markets from January 2008
statistically based tool, which is the standard deviation channel, until December 2018 was examined. It should also be noted that
a new method for determining time target zones of the price all price data used throughout the paper were obtained from
targets generated by classic price patterns is introduced here. Reuters.
The financial markets that were examined are:
Figure 2. Standard deviation channel • Equity market (S&P 500)
• Commodity market (Light crude oil)
• Currency market (EUR/USD pair)
Here, I’ve chosen seven famous classic price patterns—head
and shoulders, double bottoms and tops, triple bottoms and
tops, rectangle, ascending triangle, descending triangle, and
symmetrical triangle—and extracted them manually from daily
charts of the three aforementioned instruments. The period
of investigation from January 2008 until December 2018 was
selected, as it included a great variety of up and down trends
and nearly witnessed all market conditions where plenty of
classic price patterns could be captured.
The standard deviation channel (which is included in most
charting programs) is also alternatively referred to as the linear Types of Completed Price Patterns
regression channel. And as the name implies, it consists of two In theory, classic price patterns can’t be considered as
standard deviation bands above and below an x-period linear completed ones until an actual price breakout occurs. Further,
regression line (Figure 2). prices usually react in one of two certain ways after the
In origin, this price analysis’ tool is used for identifying the breakout. Either a pullback movement to the broken boundary
price trend and the standard deviation bands usually act as of the completed price pattern will occur before achieving the
support and resistance levels and are used for trading purposes, price target or prices will continue directly toward the target of
by the way. However, in this paper, it is used differently. the pattern.
Theoretically, the linear regression is a statistical tool that In this section, I will explain the method of determining the
is used to find a linear relationship between two variables (X time target zone for the price target of a completed price pattern
and Y), which enables one to find the value of Y (the dependent in each case. Also, it should be mentioned that this method can
variable) for each value of X (the independent variable), and in be applied to six of the seven price patterns that are examined
the case of Price and Time, we can find the value of time for the in this paper exactly in the same way; however, the head and
value of a specific price. It uses the least-squares method to plot shoulders price pattern is considered a special case, which will
a straight line through data to minimize the distances between be also explained later in this section.
the actual data and the resulting line, and to find the best
straight-line fit, the following formula is used: ŷ = b(x) + a Completed Price Patterns Without Pullback Movement After
Where [ŷ] is the predicted value, [b] is the slope, and [a] is the the Breakout
y-intercept (the value of y when x = 0). We can determine a time target zone for the price target
A distinct advantage of linear regression is that it can of a completed price pattern that isn’t followed by a pullback
forecast future price movement by simply extending the movement to the broken boundary once a price breakout occurs
resulting straight line. If we can consider that the regression by plotting the standard deviation channel on prices starting
values give the mean over the calculation period, then the from the lowest day in the last price swing before the actual
forecast is an extension of the mean. bullish breakout till the breakout day then extend the channel to
Besides, the bands of the standard deviation channel are the right (Figure 3). The time target zone is the period between
plotted to determine the approximate range of the dispersion the two vertical lines projected from interception points of
of price movements around the regression line. By default, the the price target level and the bands of the extended standard
standard deviation bands are set to two as according to the deviation channel. The opposite is true in the case of bearish
normal distribution; 95% of the observations lie within two breakouts (Figure 4).
standard deviations of the mean.
By applying the standard deviation channel tool on the last
swing of the price pattern and extend it straight forward; we
can assign a time target zone for the assigned price target of the
completed price pattern.

PAGE 28 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Figure 3. The time target zone for the price target of Figure 6. The time target zone for the price target of a
a completed bullish price pattern without a pullback completed bearish price pattern followed by a pullback
movement movement

Figure 4. The time target zone for the price target of

The special case of head and shoulders price pattern
a completed bearish price pattern without a pullback
movement The head and shoulders price pattern has a special measuring
criterion. In bullish cases, the standard deviation channel
should start from the lowest day of the pattern (the head) until
the breakout day or the last day of the pullback movement
that follows the breakout. The opposite is true in bearish cases
(Figures 7 and 8).
Figure 7. The time target zone for the price target of
bullish H&S price pattern without a pullback movement

Completed price patterns followed by pullback movements

after the breakout
If a pullback movement to the broken band of the completed
price pattern followed the actual breakout, we should add this
piece of data to the measurement. The standard deviation
channel should be started from the lowest day of the last price
swing before the breakout (or the highest day in case of bearish
breakouts) until the last day in the pullback movement after the Figure 8. The time target zone for the price target
breakout, and then extend the channel to the right (Figures 5 of bearish H&S price pattern followed by a pullback
and 6). movement

Figure 5. The time target zone for the price target of a

completed bullish price pattern followed by a pullback
movement

IFTA.ORG PAGE 29
 IFTA JOURNAL 2021 EDITION

Figure 12. Itau Unibanco holding (ITUB4.SA)—Daily

Examples From Various Financial Markets

Double bottom and double top price patterns

Figure 9. Credit Agricole bank (CAGR.PA)—Daily

From the Brazilian stock market, Figure 12 shows a daily

chart of Itau Unibanco holding, which spotted another double
top price pattern, but this time without a pullback movement.
The price target was also achieved within the determined time
target zone between November 26, 2010, and December 20,
An example from the French stock market is shown in a 2010.
daily chart of Credit Agricole bank (Figure 9). After finishing a
double bottom price pattern then a pullback movement to the Triple bottom and triple top price patterns
middle peak, the price target of the pattern was achieved later
on within the determined time target zone between August 18, Figure 13. JPY (JPY=)—Daily
2012, and September 14, 2012.
Figure 10. Jazira bank (1020.SE)—Daily

Figure 14. Light crude oil (CLc1)—Daily

Figure 10 shows a case of a double bottom pattern without a

pullback movement that occurred in a daily chart of the Jazira
bank in the Saudi stock market. The price target of the pattern
was achieved within the determined time target zone between
December 13, 2017, and January 28, 2018.
On the other hand, Figure 11 shows a completed double top
price pattern, which is followed by a pullback movement in
the Silver. Subsequently, the price hit the target of the pattern
within the determined time target zone between October 18, Figure 13 shows a case of a triple bottom price pattern
2012, and November 6, 2012. followed by a pullback movement that appeared in a daily chart
of the Japanese Yen. Again, the price target of the pattern was
Figure 11. Silver (XAG=)—Daily achieved within the determined time target zone between
November 15, 2012, and January 12, 2013.
Figure 14 shows a daily chart of the light crude oil which
illustrates a triple bottom pattern without a pullback movement
followed by a minimum target that was achieved within the
determined time target zone between January 2, 2013, and
January 25, 2013.
Here in Figure 15, a triple top price pattern appeared in a daily
chart of the Gold followed by a pullback movement afterward;
the minimum target of the completed pattern was achieved
within the determined time target zone between June 20, 2018,
and September 19, 2018.

PAGE 30 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Figure 15. Gold (XAU=)—Daily Figure 18. Deutsche Bank (DBKGn.DE)—Weekly

From the London stock market, Figure 16 shows another triple In another example from the London stock exchange, Figure
top price pattern without a pullback movement appeared in a 19 shows a bearish rectangle price pattern followed by a
daily chart of Barclays PLC. The price target of the completed pullback movement. The weekly chart of the Anglo-American
price pattern was achieved later on within the determined time stock illustrates that the price target of the completed price
target zone between October 28, 2009, and November 16, 2009. pattern was achieved later on within the determined time target
zone between February 20, 2015, and November 20, 2015.
Figure 16. Barclays PLC (BARC.L)—Daily From the French stock exchange, an example of a bearish
rectangle price pattern without a pullback movement is shown
in Figure 20. The daily chart of Danone shows that price target of
the completed price pattern was hit within the determined time
target zone between February 8, 2018, and February 28, 2018.
Figure 19. Anglo American (AAL.L)—Weekly

Bullish and bearish rectangle price patterns

The GBP pair daily chart shows a bullish continuation
rectangle price pattern followed by a pullback movement
(Figure 17). The price target of the completed price pattern
was hit subsequently within the determined time target zone
between December 10, 2013, and January 31, 2014. Figure 20. Danone (DANO.PA)—Daily
Figure 18 shows another bullish rectangle price pattern, but
this time without a pullback movement. It was spotted in a
weekly chart of Deutsche Bank which is traded in the German
stock market. Here again, the price target of the pattern was hit
within the determined time target zone between April 24, 1998,
and July 24, 1998.
Figure 17. GBP (GBP=)—Daily

Bullish and bearish ascending triangle price patterns

The weekly chart of the Canadian Dollar pair (Figure 21) shows
an example of a bullish ascending triangle pattern followed by
a pullback movement. The price target of the pattern was hit
within the determined time target zone between September 26,
2008, and January 16, 2009.

IFTA.ORG PAGE 31
 IFTA JOURNAL 2021 EDITION

Figure 21. CAD (CAD=)—Weekly

Also, from the New York stock exchange, Figure 24 shows
a daily chart of Bank of America’s stock, which captured a
bearish ascending triangle price pattern without any pullback
movements; its target was hit within the determined time
target zone between April 2, 2018, and April 19, 2018.

Bullish and bearish descending triangle price patterns

Figure 25. Ten-year U.S. treasury notes (TYc1)—Daily

Figure 22. Ten-year treasury notes (TYc1)—Daily

Another daily chart of the ten-year U.S. treasury notes (Figure

25) shows a bullish descending triangle pattern followed by a
pullback movement, and the minimum target of the pattern was
Figure 22 shows a ten-year U.S. treasury notes daily chart, achieved within the time target zone between August 26, 1998,
which spotted a bullish ascending triangle price pattern that and November 6, 1998.
wasn’t followed by a pullback movement. After finishing the Figure 26 shows a daily chart of Tadawul all shares index that
price pattern, the minimum price target was achieved within captures a bullish descending price pattern without a pullback
the determined time target zone between November 2, 2001, movement. The price target of the pattern also was achieved
and November 14, 2001. within the determined time target zone.
In Figure 23, an example of a bearish ascending triangle price In Figure 27, a bearish descending triangle price pattern,
pattern followed by a pullback movement was spotted on a daily which is followed by a pullback movement, was spotted in an
chart of General Electric Company’s stock, which is traded in the hourly chart of the Euro pair. Here again, the price target of the
New York stock exchange. The target of the completed pattern completed pattern was hit within the determined time target
was hit within the determined time target zone between March zone.
12, 2001, and March 23, 2001.
Figure 26. Tadawul all shares index (.TASI)—Daily
Figure 23. General Electric Co. (GE.N)—Daily

Figure 27. EUR (EUR=)—Hourly

Figure 24. Bank of America (BAC.N)—Daily

PAGE 32 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Figure 28. CHF (CHF=)—Daily Figure 31. Wanhua Chemical Group Co. LTD (600309.
ss)—Daily

The last example of descending triangles was a bearish one Figure 31 shows an example from the Shanghai stock
without a pullback movement that was captured in a daily chart exchange; a bearish symmetrical triangle pattern followed by
of the Swiss Franc pair (Figure 28). One more time, the price a pullback movement is spotted in a daily chart of the Wanhua
target of the pattern was hit within the determined time target Chemical Group. The price target of the pattern was hit within
zone. the determined time target zone between September 3, 2008,
and October 13, 2008.
Bullish and bearish symmetrical triangle price patterns Figure 32 shows another bearish symmetrical triangle price
pattern, but without a pullback movement. The daily chart of
Figure 29. Commercial International Bank (COMI.CA)— Alphabet INC illustrates that the price hit the target within the
Weekly
determined time target.
Figure 32. Alphabet INC (GOOGL.O)—Daily

Figure 30. Ezz Steel (ESRS.CA)—Weekly

Bullish and bearish head and shoulders price patterns

Figure 33. Orascom Development (ORHD.CA)—Hourly

From the Egyptian stock market, a case of a bullish

symmetrical triangle price pattern appeared in a weekly chart
of the Commercial International Bank (Figure 29). After a
pullback movement to the cleared boundary of the pattern, the
price trusted and hit the price target within the determined Figure 34. Deutsche boerse dax index (.GDAXI)—Daily
time target zone between December 26, 2013, and June 5, 2013.
Another case from the Egyptian stock market is in a weekly
chart of Ezz Steel’s stock, which shows a bullish symmetrical
triangle price pattern and its price target that was also achieved
within the determined time target zone (Figure 30).

IFTA.ORG PAGE 33
 IFTA JOURNAL 2021 EDITION

Figure 37. Frequency distribution of successful price

An hourly chart of Orascom Development Egypt shows a patterns that appeared in the three instruments across
bullish head and shoulders price pattern followed by a pullback the investigation period
movement and its target that was hit within the determined
time target zone between November 22, 2018 at 1:00 pm, and
December 25, 2018 at noon (Figure 33).
Figure 34 shows another bullish head and shoulders price
pattern but with no pullback movement. The daily chart of the
main index of the Germain stock market illustrates that the
price target of the completed pattern was achieved within the
determined time target zone.
Figure 35. Egyptian Financial and Industrial SAE (EFIC.
CA)—Weekly

Figure 38 shows that 92% of double tops and bottoms

patterns succeeded in achieving the price target of the pattern
within the determined time target zone, while only 8% failed;
95% of triple tops and bottoms price patterns succeeded while
only 5% failed; 100% of rectangle price patterns succeeded; 84%
of ascending triangle price patterns succeeded while 16% failed;
Figure 36. EGX30 (.EGX30)—Daily 91% of descending triangle price patterns succeeded while only
9% failed; 71% of symmetrical price patterns succeeded while
only 29% failed; and finally, 84% of the H&S price patterns
succeeded in achieving the price target within the determined
time target zone while only 16% failed.
Figure 38. The success rate of achieving the price target
within the determined time target zone for all patterns

Two more examples of bearish head and shoulders price

pattern from the Egyptian stock market appear in Figures
35 and 36. The weekly chart of the Egyptian Financial and
Industrial stock shows an example of the pattern followed
by a pullback movement while the daily chart of the main
index shows another example without a pullback movement.
In the two cases, the price targets were achieved within the
determined time target zone.

Statistical Analysis of Time Target

Zones of Price Targets of Classic
Price Patterns
In a visual inspection of the daily data of the S&P 500 index,
the Light crude oil and the EUR/USD pair from January 1, 2008,
until December 31, 2018, disclosed 38 double top or bottom
patterns, 21 triple top or bottom patterns, 27 rectangle patterns,
19 ascending triangle patterns, 23 descending triangle patterns,
28 symmetrical triangle patterns, and 49 head and shoulders
patterns that achieved their price targets (Figure 37).

PAGE 34 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Figure 41. Frequency distribution of successful price

Figure 39 shows that most of the bullish patterns have patterns that followed by pullback movements
succeeded in hitting their price target within the determined that appeared in the three instruments across the
time target zone more than triple the failure times except for investigation period
the bullish symmetrical triangle pattern, which recorded a
moderate success rate.
Figure 39. The success rate of achieving the price target
within the determined time target zone for bullish
patterns

Figure 42. The success rate of achieving the price target

within the determined time target zone for bullish
patterns with pullback movements

Figure 40. The success rate of achieving the price target

within the determined time target zone for bearish
patterns

Figure 42 shows that again, except for the bullish symmetrical

triangle patterns followed by pullback movements, which
recorded a 1:1 success rate, five patterns recorded the full mark,
and the rest recorded sufficiently high success rate.
Figure 43 shows superiorly high success rates for bearish
patterns followed by pullback movements, including higher
rates for the bearish symmetrical triangle and bearish
ascending triangle patterns than the bullish ones.

On the other hand, Figure 40 reflects the superiority of the Figure 43. The success rate of achieving the price target
within the determined time target zone for bearish
bearish patterns with four patterns recording the full mark and
patterns with pullback movements
the rest recording sufficiently high success rates.
Figure 41 shows the number of patterns followed by a pullback
movement before achieving the price target out of the total
patterns that appeared.

IFTA.ORG PAGE 35
 IFTA JOURNAL 2021 EDITION

Figure 44. Frequency distribution of successful price Figure 47. Percentages achieving the price target within
patterns that weren’t followed by pullback movements the first or second half of the determined time target
which appeared in the three instruments across the zone for bullish patterns
investigation period

Figures 47 and 48 show that in general, price targets of the

Figure 44 shows the number of successful patterns that bearish patterns are achieved most often within the first half
achieved their price target directly after the breakout without of their time target zones, despite the bullish ascending and
pullback movements out of the total patterns that appeared. descending triangle patterns proving better performance in
that matter.
Figure 45. The success rate of achieving the price target
within the determined time target zone for bullish Figure 48. Percentages achieving the price target within
patterns without pullback movements the first or second half of the determined time target
zone for bearish patterns

Figures 45 and 46 show that the bearish patterns that weren’t

followed by a pullback movement recorded higher success rates
than the bullish ones except for the symmetrical triangle and
the head and shoulders patterns. Summary
Identifying an expected return (or—as in technical analysis’
Figure 46. The success rate of achieving the price target terminology—a price target) is an advantage in the financial
within the determined time target zone for bearish
markets environment, however definitely, if it is accompanied
patterns without pullback movements
by a time target to be achieved in, it will be superior.
This paper introduced an approach that tried to add more
value to seven of the classic price patterns through the ability
to identify time target zones for their traditional price targets
by applying linear regression on the last price swing in the
price pattern.
A variety of examples from various financial markets on
multiple timeframes and throughout multiple decades were
included in this paper as proof of the validity and applicability of
the thesis throughout the age of the classic price patterns.
Finally, like any tool or approach, our approach also is not
infallible, despite the result of the statistical analysis of the
examined data reflecting satisfying overall success rates.
Moreover, the bearish patterns seemed to have the upper hand,

PAGE 36 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

especially in achieving the targets within the first half of the

determined time target zone.

References
Aronson, David R., Evidence-Based Technical Analysis. New Jersey: John
Wiley & Sons, 2007.
Bulkowski, Thomas N., Encyclopedia of Chart Patterns. New York, United
States: Wiley, 2005.
—. Trading Classic Chart Patterns. United States: Wiley, 2002.
Edwards, Robert D., John Magee and W.H.C. Bassetti, Technical Analysis
of Stock Trends, 9th ed. CRC Press, 2007.
Kaufman, Perry J., Trading Systems and Methods, 5th ed. Hoboken, New
Jersey: John Wiley & Sons, 2013.
Kirkpatrick, Charles D. and Julie R. Dahlquist, Technical Analysis: The
Complete Resource for Financial Market Technicians. FT Press, 2006.
Murphy, John J., Technical Analysis of the Financial Markets. Welwyn
Garden City, United Kingdom: NYIF, 1998.
Pring, Martin J., Technical Analysis Explained, Fifth Edition. United
States: MCGRAW-HILL, 2014.

Notes
1 Murphy, John J. Technical Analysis of the Financial Markets. Welwyn
Garden City, United Kingdom: NYIF, 1998, P. 108-110.
2 Kirkpatrick, Charles D. and Julie R. Dahlquist, Technical Analysis: The

Complete Resource for Financial Market Technicians. FT Press, 2006,

P. 302-303.
3 Pring, Martin J., Technical Analysis Explained, 5th ed. United States:

MCGRAW-HILL, 2014, P. 65.

4 Kaufman, Perry J., Trading Systems and Methods, 5th ed. Hoboken,

New Jersey: John Wiley & Sons, 2013, P. 236.

5 Kaufman, Perry J., Trading Systems and Methods, 5th ed. Hoboken,

New Jersey: John Wiley & Sons, 2013, P. 238-239.

6 Kaufman, Perry J., Trading Systems and Methods, 5th ed. Hoboken,

New Jersey: John Wiley & Sons, 2013, P. 250.

7 Aronson, David R., Evidence-Based Technical Analysis. New Jersey:

John Wiley & Sons, 2007, P. 212.

8 Edwards, Robert D., John Magee and W.H.C. Bassetti, Technical

Analysis of Stock Trends, 9th ed. CRC Press, 2007, P. 62.

9 Murphy, John J. Technical Analysis of the Financial Markets, Welwyn

Garden City, United Kingdom: NYIF, 1998, P. 106.

Software and Data

Data and charts provided by Reuters.

IFTA.ORG PAGE 37
 IFTA JOURNAL 2021 EDITION

Ron Albert Marcelino Acoba, CFTe, CMT, MFTA

Emerging Currencies as Equity Earthquake 39 Guatemala Street, Loyola Grand Villa
Barangay Pansol
Indicator Quezon City, Metro Manila 1108
Philippines
By Ron Albert Marcelino Acoba, CFTe, CMT, MFTA
+639175814766
[email protected]

Abstract Historical Anecdotal Precedents

Investing in the equity markets traditionally entails
institutional analysts and investors alike to primarily focus Asian financial crisis
on corporate earnings and the health of the economy by The epicenter of the Asian financial crisis that crumbled much
looking at macro indicators such as GDP growth, inflation, an of East Asia’s economy was in Thailand. For most of the 1980s
unemployment rate, among others, along with the price and and 1990s, Thailand, along with South Korea, Singapore, Taiwan,
volume actions of the general market and individual issues. and Hong Kong, was in a boom cycle. The group was actually
Not much attention, however, is given to the impact of adverse termed as “Tiger economies” due to the high GDP growth that
movements in currencies in relation to equities. Unbeknownst they were seeing in the period. Thailand’s economy for its part at
to many retail investors, currency volatility is a major catalyst one point grew by a high of 11.2% in the 1990s. The high growth
of hot money flight. The aim of this paper is to demonstrate that the region, particularly Thailand, was seeing attracted a
that the movements in emerging currencies can be used as a lot of foreign funds. Cheap interest rates of dollar denominated
seismograph to forecast short-run equity movements and as borrowings, false security by corporate borrowers due to the
cues to actively defend against market aftershocks. Thai government’s encouragement or “guarantee,” and a fixed
exchange rate regime eventually led to over borrowing. There
Acknowledgements came a point where funds were being channeled to negative NPV
I would like to thank my wife Charlene, who is also my ventures (Corsetti and others 1998).
business partner in Trading Edge, for continually inspiring me An unprecedented amount of capital inflows, together with a
and motivating me in advancing our research in the financial drop in exports, widened the country’s current account deficit
markets particularly in the field of technical analysis. I would (Aghevli 1999). What made matters worse was that a good
also like to thank my family, parents, and brother for their portion of the capital inflows came in as short-term borrowings
constant support in our professional endeavor. Last but not (Carson and Clark 2013). This situation eventually sparked an
the least, I would like to express my gratitude to my students, external attack by shorters on the Thai Baht (Dumlao 2017).
subscribers, Trading Edge’s corporate clients, and media It also did not help that the government burnt a good amount
partners for their confidence and trust in our work. of their dollar reserves in their attempt to keep and protect
the peg. The 25 Baht to $1 peg was shortly uncorked to 30
Introduction Baht to $1 on July 2, 1997. This event led to a series of currency
devaluations that griped East Asia. In the first seven months
Seismograph after the central bank of Thailand unpegged their exchange
A seismograph or a seismometer, first invented in the 2nd rate, the Baht fell to a low of 56.75 against the USD. Similarly,
century in China, is an instrument that measures and records the Korean Won fell to a low of 1,995 against 1 USD from only
the movements in the earth or “seismic waves” caused by 888 at the end of June 1997. The Indonesian Rupiah suffered a far
an earthquake, volcanic activity, explosion, or other earth- worse blow as it weakened to a low of 15,650 against 1 USD from
moving phenomena. A seismograph produces a record of earth only 2,431.50. The Malaysian Ringgit and the Philippine Peso
movements onto a screen or a paper printout. This is called a likewise suffered more or less the same fate as their respective
seismogram or, in terms of the financial markets, a price chart. currencies melted to 4.88 from 2.52 and 46.80 from 26.38
History is rich with notable earthquakes—from the deadliest against the greenback.
ever recorded in Shaanxi, China, in 1556, to Kobe in Japan As investors dropped the Asian currency denominated assets
in 1995, to Aceh, Indonesia, in 2004, and to Haiti in 2010. In like hot stone fresh from a spewing volcano, the Asian equity
a similar fashion, the financial markets are likewise rich in markets likewise faced a similar demise. The Thailand Stock
notable currency events that placed a lot of economies and Market (SET50) fell from a high of 535 in July 1997 to a low of 134
investors alike to their knees. This paper first looks at the in the next 18 months. The Korea Stock Exchange Composite
history of different currency crises in the emerging markets Index (KOSPI) fell from a high of 782 to a low of 277 in the same
and assesses their consequent impact on their respective equity period. Indonesia’s Jakarta Stock Price Index (JCI) fell from a
market. We can then use the quantitative observations of the high of 742 to a low of 255. Malaysia’s Kuala Lumpur Composite
past to actively defend our portfolio from aftershocks in the Index (KLCI) declined to a low of 261 from a high of 1,085 while
equity markets should they occur later on. the Philippine Stock Exchange Index (PHISIX) slumped to a low
of 1,075 from a high of 2,815 in the 18 months after July 1997.

PAGE 38 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Table 1. THBUSD versus SET 50 (July 1, 1997, to December

30, 1998) bribes and kickbacks to different politicians while the rest
were allegedly embezzled for personal the use of Najib and
Low (Wright 2015). Swiss authorities said that as much as $7
billion flowed from 1MDB and its units (Adam and others 2018).
The news of embezzlement, corruption, money laundering,
and betrayal of public trust, among others, by no less than the
prime minister of Malaysia, cast an ash cloud on the country’s
economy and political leadership.
Table 2. KRWUSD versus KOSPI (October 17, 1997, to April Prior to the 1MDB scandal, the Malaysian Ringgit was already
19, 1999) reeling because of the 45% slide in the global prices of crude
(WTIC) from its June 2014 high to the end of June 2015. Crude
was Malaysia’s number 1 export and a slip in price naturally
meant lesser dollars flowing into the government’s coffers
(Y-Sing 2014). The scandal that came into light in July all the
more gave foreign investors the reason to flee the country. This
was then aggravated further when China devalued the Yuan in
August 2015 in their attempt to shore up their exports (Farrer
2015). From July 1, 2015, the US dollar to Malaysian Ringgit
Table 3. IDRUSD versus JCI (July 9, 1997, to January 11,
1999) exchange rate zoomed from 3.78 to a high 4.48 inside just 3
months. In the same time period, the Kuala Lumpur Composite
Index fell from 1,706 to a low of 1,503.
Table 6. MYRUSD versus KLCI (July 1, 2015, to January 2,
2017)

Table 4. MYRUSD versus KLCI (July 10, 1997, to January

11, 1999)

Indian Rupee Slide

The US dollar to Indian Rupee exchange rate of the USDINR
pair had surged 16.68% to a high of 74.48 in October 2018 from
just 63.83 at the beginning of the year. The steep slide of the
Rupee can be attributed to a confluence of global economic
Table 5. PHPUSD versus PHISIX (July 10, 1997, to January developments. Firstly, the U.S. Federal Reserve embarked
11, 1999) on a more aggressive monetary policy. It continued to raise
its benchmark interest from 1.5% to 1.75% in March 2018
after already raising it three times in the prior year. Robust
employment figures and GDP growth led the Fed to hike its rates
anew to 2% in June while signaling that it would do so for a total
of four times in all of 2018. The Fed’s actions sparked a massive
capital outflow from emerging markets as U.S. yields became
more attractive. Consequently, the Fed’s policy also lessened
1Malaysia Development Berhad (1MDB) Scandal the availability of the USD in the global system, which further
The 1MDB scandal is an incident that placed a spotlight on pushed its value and likewise placed downward pressure on
Malaysia’s corruption-laden government dealings. 1MDB was emerging currencies like the Rupee (Sengupta 2018).
a government-run investment company with the mission to Another event that contributed to the Rupee’s slide was the
promote the country’s long term development and foreign 28.3% surge in the global prices of crude as measured in the
capital investments. Its interests were mainly centered on West Texas Intermediate Crude (WTIC). WTIC jumped from just
energy, tourism, real estate, and agriculture. In July 2015, $60.07 per barrel at the start of 2018 to a high of $77.07 per
then Prime Minister Najib Razak, who was the advisory board barrel in October. A decrease in daily production by OPEC and
chairman, was accused of siphoning roughly $700 million from Russia plus the United States’s sanctions on Iran curbed the
the company to his personal accounts (Wright and Clark 2015). supply of crude, which therefore translated into an increase in
The U.S. DOJ claimed that the in-crowd of the prime minister its price. With India importing about 70% of its crude demand,
led by businessman Jho Low channeled money from 1MDB to the increase in its price only worsened India’s current account
various shell companies. Some of the funds were used to pay deficit. In addition, the “trade war” between the United States

IFTA.ORG PAGE 39
 IFTA JOURNAL 2021 EDITION

Table 8. TRYUSD versus BIST 100 Index (January 1, 2018,

and China plus the latter’s currency devaluation further to January 1, 2019)
triggered more emerging market-to-develop markets flows
(Singh 2019). In the period where the USDINR rose by 16.68% to
74.48 in October 2018 from 63.83 at the start of January, India’s
BSE SENSEX still managed to edge higher by 2.07% to 34,760.
This came despite a strong GDP growth of 7.1% to 8.2% in its
most recent quarters. It would have fared much higher if only
the Rupee had been stable.
Table 7. INRUSD versus BSE SENSEX (January 1, 2018, to 2015 Ghost Month: The Yuan Devaluation
January 1, 2019)
The global financial markets, particularly the Asian emerging
markets, were visited early by the “hungry ghosts” as China
enacted a series of unexpected devaluations of its currency,
the Yuan, against the U.S. dollar. On August 11, 2015, exactly
three days before the beginning of the Chinese Ghost Month,
the Chinese central bank weakened the Renminbi by 1.86%
to 6.2298 per U.S. dollar (Mo 2015). A day later, a huge tidal
wave of selling continued to grip the financial markets as the
Turkey Currency Crisis People’s Bank of China pushed their currency down by another
Turkey’s currency crisis bore as a result of the country’s 1.62% against the greenback (Riley 2015). It did not actually
appetite for foreign-denominated debt. In the recent years, stop there, as the bank, for a third day in a row, softened it
Turkey’s economy has been growing at a very robust clip, even further to 6.3975 per U.S. dollar (Inman 2015). The central
rising by a high of 7.5% for the entire 2017. Much of its growth, bank’s actions came following an 8.3% slide in exports in July
however, was fueled by high levels of government expenditure from a year ago (Vaswani 2015). Many believe that the bank’s
funded by foreign-denominated debt (Lee 2018). As of the end moves were intended to shore up the country’s international
of 2017, Turkey’s gross external debt, which includes public and trade. According to the PBOC, however, the “adjustment in
private loans, reached $453.20 billion. With roughly 40% of the the fixing rate formation mechanism was aimed at correcting
said amount, or ~$180 billion, set to mature in 2018, plus the the disparity between the fixing rate and the spot rate of the
fact that the country only has a gross currency reserve of $85 Chinese Yuan against the U.S. dollar” since the Yuan, being
billion, investors started to become wary of Turkey’s capability pegged with the USD, had been steadily rising in value along
to service its maturing debt (Holland 2018). To add fuel to the with the greenback on the back of the planned monetary
fire, the USD had likewise started to strengthen on the back tightening in the United States. Additionally, PBOC’s actions
of the U.S. Federal Reserve’s tightening policies. Additionally, were also meant as a bid to join the IMF’s Special Drawing Rights
the Central Bank of the Republic of Turkey’s decision to keep basket (Chin 2015).
its interest rates steady at 8% even at the face of a double- Nonetheless, shockwaves were sent not just in the currency
digit inflation rate added more pressure on the currency. It markets but likewise in the equities and commodities spaces
was actually Turkish President Recep Tayyip Erdogan who following the surprise move from the Chinese. With the USDCNY
prevented the central bank from increasing their interest rates. exchange rate rising from 6.2093 in August 10, 2015, to a high
He most famously, or infamously, rather stated that “interest of 6.4485 in the next three days, capital outflows from China
rates are the mother and father of all evil” (Baccardax 2018). and even from the rest of Asia accelerated as a consequence. In
The president of Turkey’s interference with the central bank the following two weeks, the Shanghai Composite Index fell by
ran contrary to the notion that central banks should be left 27.44% from 3,928 on August 10 to a low of 2,850.
independent. Furthermore, the economic sanction imposed by
the United States to Turkey by doubling its tariffs on Turkish Table 9. CNYUSD versus SSE Composite (August 10, 2015,
to February 10, 2017)
steel and aluminum placed more pressure on the country’s
already ballooning current account deficit (Harper 2018).
As expected, investors let go of Turkish financial assets
following the steep slide in the Turkish Lira. The above factors
resulted to a sharp depreciation of the Lira against the U.S.
dollar, with the USDTRY exchange rising to a high of 7.083 in
August 2018 from only 3.791 at the end of December 2017. This
of course cascaded into a 26.6% drop in the Borsa Istanbul 100
Index. Literature
John Murphy, CMT, a living legend in the field of technical
analysis, noted the concept of intermarket analysis in his book
Trading with Intermarket Analysis (2012). Here, he said that
intermarket analysis involves the examination of different
asset classes, bonds, stocks, commodities, and currencies. The

PAGE 40 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

premise of intermarket analysis is that these asset classes are the exchange rates impact the competitiveness of firms, as
related. This means that what happens in one market also has it affect the firms’ cost of capital and earnings which in turn
an impact on the other asset classes. In his book Intermarket get reflected on its share price. They added that on a macro
Analysis: Profiting from Global Market Relationships, he noted level, the impact of exchange rate movements on the equity
that “a falling currency usually gives a boost to commodity market would likewise be hinged on the “degree of openness of
prices quoted in that currency. This boost in commodity prices the economy and the degree of trade imbalance.” Christopher
reawakens inflation fears and puts pressure on central bankers Ma and G. Wenchi Kao (1990) found that on a macro level, an
to raise interest rates, which has a negative impact on the appreciation of the domestic currency negatively affects the
stock market” (Murphy 2004). He actually first introduced domestic stock market of an export-oriented economy, while a
the concept of intermarket analysis in his book Intermarket depreciation of the local currency positively affects the equity
Analysis: Trading Strategies for the Global Stock, Bond, market of an import-oriented one.
Commodity, and Currency Markets. Here, he mentioned that a
“rising dollar will eventually push inflation and interest rates Methodology
lower, which is bullish for (US) stocks and a falling dollar will
eventually push (US) stock prices lower because of the rise in Association Analysis
inflation and interest rates. However, it is an oversimplification For association analysis, Pearson Correlation Analysis was
that a rising dollar is always bullish for (US) stocks and a falling used to find the correlation of 10-year historical prices of
dollar is always bearish for (US) equities” (Murphy 1991). It is emerging market indices against their respective currency
the aim of this paper to see whether this premise also applies (historical data of sample equity indices, their respective
in the emerging market economies, specifically in East Asia. Of currency exchange, and the US Dollar Index were sourced
course, other dimensions, like the Asian economy’s exposure to from Investing.com and TradingEconomics.com). This type of
dollar denominated debt, exposure to hot capital flows, and the correlation analysis is used to determine the degree of linear
account of imports and exports in relation to their GDP, also play relationship between two quantitative variables that are
important roles. assumed normally distributed. Essentially, three questions will
In a study done by Clive Granger, Bwo-nung Huang, and Chin be answered by the corresponding Pearson’s coefficient: (1) Is
Wei Yang (1998), evidence was found of a bivariate causality there a linear relationship? (2) How strong is it? (3) What is the
between stock prices and exchange rates. Using the Asian direction of the linear relationship?
flu data of the 1990s, they found that in at least South Korea Sample Pearson correlation coefficient is computed as
and Thailand, exchange rates lead stock prices with a positive s XY
correlation and that the “inclusion of exchange rate variations r=
is found to have improved the predictable portion of stock price s X sY
changes of the eight Asian economies.” They also found, though,
that at times, stock prices lead the changes in the exchange where the numerator estimates the covariance between the
rates. They further concluded that with the opening of Asian two variables (say X and Y), and the denominator is the product
economies to trade and global financing, capital movement between the two sample standard deviations.
into and out of the Asian economies can be both beneficial and The sample Pearson correlation coefficient, r, takes on values
detrimental, as in the case of the Asian financial crisis, where from -1 to +1. A value very near -1 indicates an almost perfect
a lot of capital flew out of the region. Abdul Qayyum and A.R. inverse (or indirect) linear relationship between two variables,
Kemal (2006) also found that volatility in the exchange rate while a correlation coefficient very near +1 indicates almost
has a spill over in the Pakistani stock market and vice versa. perfect direct linear relationship. On the other hand, a value
In the same way, Dr. Alok Kumar Mishra, A.K. Swain, and D.K. very near zero implies absence of linear relationship, but this
Malhotra (2007) found bi-directional volatility spillovers does not mean absence of another form of relationship (other
between the Indian equity market and its exchange rate. than linear).
Meanwhile, Usman Umer, Guven Sevil, and Serap Kamisli (2015) As a rule of thumb, the qualitative interpretation for the
determined that the “comovements between exchange rates correlation coefficient is given below. If the coefficient is
and stock prices become stronger during the crises time, and the negative, the variables are moving at opposite directions (as
direction of causality originates from stock prices to exchange one decreases, the other one increases or vice versa) but if the
rates during the tranquil period; and from exchange rates to coefficient is positive, the variables are moving in the same
stock prices during crisis.” direction (as one decreases [or increases], the other one also
Additionally, Valentina Bruno and Hyun Song Shin (2018) decreases [or increases]).
found that “non-financial firms that exploit favorable global 0: No association
financing conditions to issue bonds and build cash balances 0.01 – 0.20: Very weak association
are also those whose share price is most vulnerable to local 0.21 – 0.40: Weak association
currency depreciation.” They further found that while currency 0.41 – 0.60: Moderate association
depreciation would be favorable for exporting firms, the 0.61 – 0.80: Strong association
impact of the financial burden from their foreign-denominated 0.81 – 0.99: Very strong association
debt outweighs their supposed competitiveness. According 1: Perfect association
to Rudiger Dornbusch and Stanley Fischer (1980), changes in

IFTA.ORG PAGE 41
 IFTA JOURNAL 2021 EDITION

The association tables note the year group used in the In here, we tested for the stationarity of the levels (actual
analysis. Year group 2017–2018 used all observations within value) and the first-differences (current minus previous). First-
these years, as is the case for year group 2011–2018. Take note difference is a useful concept in time-series modelling since it
that for Thailand and India, only up until 2011 is available, so shows the change between successive values.
the year groups for these countries are the recent two years
and the recent eight years. For the other countries, year 2009 is Dependence of Stock Market: Regression Models of
available, so the year groups are 2017–2018 (recent two years) Stock Market
and 2009–2018 (recent 10 years). Four linear regression models for change in the equity
For each year group, the stock market of the of a sample Asian index (dependent variable) are used. Two models used only the
emerging market represented by its equity index is correlated individual exchange rate as predictor at varying year groups and
to the respective local currency exchange rate and USD Index another two models used both the exchange rate and the USD
(reciprocal) resulting to four coefficients stated under the 3rd Index (reciprocal) as predictors, also at varying year groups.
column. Qualitative interpretation for each coefficient is also Each table under a certain year group lists the coefficients
given under the 4th column. The last column lists the p-value (impact or effect) of the predictor to the change in the equity
associated with the estimated correlation coefficient. p-values index. Accompanying each coefficient is its p-value. Stated in the
help us decide for the statistical significance of the estimate. last row is the coefficient of determinations R-squared which
From this point onward, we reject the null hypothesis (Ho) of tells us the percentage of variation in the dependent variable
the appropriate test if the p-value is small (e.g., less than 0.10). that can be explained by the predictors. Higher R-squared
Otherwise, Ho is accepted. Rejecting the Ho means that the means better predictive ability of the model. If the coefficient
conclusion will support the alternative hypothesis (Ha). For and p-value are blank, it means that the corresponding
Pearson Correlation Analysis, the set of hypotheses are: predictor is not included in that model. Take note that predictors
Ho: The two variables are not associated. are dictated by the lag order selected in previous analyses.
Ha: The two variables are associated. For example, the THB/USD exchange rate in 2017–2018 has
lag order 1, so the lag of the first-differenced THB/USD exchange
Significant results (Ho is rejected) are noted by two asterisks rate denoted by L1D is included in the model below. D refers to
(**) printed beside the p-value while nonsignificant results (Ho the first-difference while L refers to the lag, and the number
is accepted) are noted by (ns). attached to it is the order.

Dependence of Stock Market: Selection of Lag MSCI Emerging Market Index Versus MSCI EM
Order Currency Index Test
Lag refers to the interval between present and past values On top of the establishment of correlation and causation
and is vital when dealing with time series data since past values between samples of individual emerging market currencies to
are usually correlated to present values. Before modelling, it is their respective equity markets (Baht vs. SET, Rupee vs. BSE
important to know which lag has the highest contribution to the SENSEX, Lira vs. Borsa Istanbul, Ringgit vs. Bursa Malaysia,
behavior of the time series data. The criteria that we will use to Rupiah vs. JCI), and domestic currency along with the reciprocal
determine the number of lags in the model are final prediction of the US Dollar Index against the respective equity market for
error (FPE), Akaike’s information criterion (AIC), Schwarz’s the periods 2017–2018 and 2009–2018, the paper will also look
Bayesian information criterion (SBIC), and Hannan-Quinn for the correlation and causation between the MSCI Emerging
information criterion (HQIC). Smaller values for these criteria Markets Currency Index and the MSCI Emerging Markets Index.
mean better modelling capability. If the lag order is suggested It is likewise important to subject the MSCI Emerging Markets
by most of these criteria, that lag order is used. If exactly two of Index against its currency index because of the widespread
these criteria suggested two different lag orders, the lowest order use of both indices as a benchmark of global funds. As per
is chosen due to issue of parsimony and ease of interpretation. the MSCI.com website, the MSCI Emerging Market Index was
Parsimony refers to the ability of the model to explain the launched in 1988. Back then, it only consisted of 10 countries
dependent variable given minimal number of predictors. representing less than 1% of global market capitalization. Since
then, the index has grown to include 24 economies, namely
Dependence of Stock Market: Association Tests Brazil, Chile, China, Colombia, Czech Republic, Egypt, Greece,
For time-series modeling to be applicable, the time-series Hungary, India, Indonesia, Korea, Malaysia, Mexico, Pakistan,
must be stationary. Through time, the behavior of the variable Peru, Philippines, Poland, Russia, Qatar, South Africa, Taiwan,
must not show any trend (increasing or decreasing) or Thailand, Turkey, and the UAE. As of September 2018, China
fluctuations that resemble a pattern. Such time-series must be has the heaviest weight, accounting for 30.99% of the index.
detrended or must be transformed into a stationary process. South Korea and Taiwan account for 14.88% and 12.29%,
Augmented Dickey-Fuller Tests can check whether the time- respectively, in the index. The index captures both large and
series data is stationary or not by hypothesizing about the mid-cap representations across the said countries and also
unit-root. If a variable has a unit-root, then the variable follows a covers approximately 85% of the free-float adjusted market
random walk or a behavior with an unpredictable pattern. capitalization in each country.
Ho: The time-series data has a unit-root. (Not stationary) Meanwhile, the MSCI EM Currency Index includes currencies
Ha: The time-series data is a stationary process. from 25 emerging market countries. Currencies from Asian

PAGE 42 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

emerging markets account for a little over half of the weight of Dependence of Bangkok SET50 Index: Selection of Lag Order
the index at 50.2%. On the other end, currencies from emerging
Europe and the Middle East (EMEA) and emerging Latin America Table 11. Selected Lag Orders and LR Test P-value for
SET50 to THBUSD and US Dollar Index
comprise about a quarter each of the weight of the index. For
emerging Asia, the weights of the Chinese Yuan and the Korean Selected LR Test
Won are 14.5% and 13%, respectively. Additionally, the Brazilian Year Group Variable Lag Order Suggested by: p-value
Real has a 16.9% weight under the emerging Latin America. Bangkok
Theoretically, strength in the EM Currency Index should 1 all four <.001 **
SET50 Index
translate to a consequent rise in the equity space as measured in
THB/USD
the MSCI Emerging Market Index and vice versa. Using the daily 2017–2018 1 all four <.001 **
Exchange Rate
data from January 12, 2015, to April 1, 2019, we shall establish
a correlation and causation via regression analysis between USD Index
1 HQIC, SBIC <.001 **
(reciprocal)
the two indices. Additionally, we will also look to quantify the
resulting decline measured in the following month, two and Bangkok
1 HQIC, SBIC <.001 **
three months in the equity index for every 2% break in the SET50 Index
currency index. A “break” in the currency index is defined as THB/USD
a fall below an established swing low in its daily bar chart. The 2011–2018 1 HQIC, SBIC <.001 **
Exchange Rate
same will also be measured for a 2% decline from an established USD Index FPE, AIC,
swing high or peak. These technical breaks in the currency 2 0.035 **
(reciprocal) HQIC
index shall be deemed as “signals” for a possible drop in the
emerging equity index. For the most recent two years, all variables are somehow
dictated by the most recent previous value (lag of 1). The
Results behavior of the Bangkok SET50 Index and THB/USD exchange
rate is consistent even for the most recent eight years. However,
Results for Thailand the USD Index (reciprocal) is shown to be affected generally by
the two previous values.
Association of Bangkok SET50 Index to THB/USD Exchange Note that these lags will be considered in the succeeding
Rate and USD Index (reciprocal) regression models.
Table 10. Correlation coefficients and P-values of SET50 Dependence of Bangkok SET50 Index: Stationary Tests
to THBUSD and US Dollar Index
Table 12. Results of Dickey-Fuller Test for Bangkok
Association of Pearson’s
Year Stock Market Correlation SET50 Index to THBUSD and USD Index
Group to: Coefficient Interpretation p-value Dickey-Fuller Test
THB/USD very weak & 0.002 p-value
0.1393
2017– Exchange Rate direct ** Year At the First-
2018 USD Index very weak & Group Variable level Difference Remark
0.0541 0.232 ns
(reciprocal) direct First-
Bangkok
THB/USD 0.609 ns <.001 ** differenced data
0.2400 weak & direct <.001 ** SET50 Index is stationary.
2011– Exchange Rate
First-
2018 USD Index very weak & 2017– THB/USD
0.0322 0.161 ns 0.972 ns <.001 ** differenced data
(reciprocal) direct 2018 Exchange Rate is stationary.
First-
USD Index
The Bangkok SET50 Index had a very weak (r = 0.1393) direct 0.881 ns <.001 ** differenced data
(reciprocal) is stationary.
linear association to THB/USD Exchange rate for the years 2017
to 2018 but shared stronger (r = 0.2400) direct linear association Bangkok
First-
for the past eight years. Both were found to be statistically 0.854 ns <.001 ** differenced data
SET50 Index is stationary.
significant. On the other hand, its association to the USD
Index (reciprocal), aside from being very weak, was also not First-
2011– THB/USD
0.816 ns <.001 ** differenced data
significant. This implies that the Thailand stock market does not 2018 Exchange Rate is stationary.
share linear association with the reciprocal of the USD Index.
First-
All associations were positive or direct, which means that when USD Index
0.883 ns <.001 ** differenced data
THB/USD exchange rate or USD index (reciprocal) increases (reciprocal) is stationary.
(or decreases), the Bangkok SET50 Index also increases (or
decreases). For all variables in either year groups, first-differences follow
a stationary process. Since first-differences follow a stationary
process, then the models for all countries will predict the change

IFTA.ORG PAGE 43
 IFTA JOURNAL 2021 EDITION

Table 14. Regression results for Bangkok SET50 Index to

in the equity index and not the actual observed stock market. All THBUSD and USD Index
interpretations should focus in the change in the equity index.
2017–2018 2011–2018
However, take note that even if interpretations are directed to
“change” in the equity index, mathematical characteristics of Predictor Coefficient p-value Coefficient p-value
time series model can extend the interpretation at the levels (L1D). Bangkok
(actual or non-differenced values). In other words, since all 0.00007 0.999 ns 0.002 0.934 ns
SET50 Index
variables across different timeframes that were tested showed
(D). THB/USD
that their respective time series is a stationary process, this Exchange Rate
17433.39 0.013 ** 37893.83 <.001 **
essentially means that the change in the Bangkok SET50 Index
is somewhat predictable by the movements in the THB/USD (L1D). THB/USD
9289.26 0.153 ns 4721.59 0.226 ns
exchange rate and the USD Index (reciprocal) and is not due to Exchange Rate
“random walk.” (D). USD Index
-4478.54 0.702 ns -21724.50 0.004 **
(reciprocal)
Dependence of Bangkok SET50: Regression Models of Bangkok (L1D). USD Index
SET50 1956.53 0.879 ns -11770.09 0.058 **
(reciprocal)

Table 13. Regression results for Bangkok SET50 Index to (L2D). USD Index
- - -2001.86 0.727 ns
THBUSD (reciprocal)
R-squared 2.62% 6.58%
2017–2018 2011–2018
Predictor Coefficient p-value Coefficient p-value
Incorporating the USD Index (reciprocal) into the model
switched the coefficient of previous change in the Bangkok
(L1D). Bangkok SET50 Index from negative to positive. With a small effect, the
-0.0003 0.996 ns 0.002 0.949 ns
SET50 Index test suggested that changes in the Bangkok SET50 Index are not
(D).THB/USD significantly dependent to previous changes. With or without
16455.6 0.010 ** 32884.91 <.001 **
Exchange Rate the USD Index (reciprocal), the THB/USD exchange rate still has
(L1D).THB/USD significant direct effect to the Bangkok SET50 Index, especially
9694.75 0.102 ns 1671.32 0.626 ns the current (D) change in the THB/USD exchange rate.
Exchange Rate
The models above suggest that the USD Index (reciprocal)
R-squared 2.59% 5.78%
appears to have long-term effect to Thailand stock market. The
Clearly, the THB/USD exchange rates are significantly USD Index (reciprocal) is not significant in the 2017–2018 year
positively associated to the Bangkok SET50 Index (positive group but is significant in the 2011√2018 year group up to the
coefficients) which implied that an increase in THB/USD first lag difference (L1D). The model of the longer year group also
exchange rates also increases the Bangkok SET50 Index. For suggests that the USD Index (reciprocal) has indirect effect to
2017–2018, one unit increase in the previous (lag 1) change the Bangkok SET50 Index, as seen in its negative coefficients.
(difference) in the Bangkok SET50 Index caused a decrease Specifically, an increase in the difference between successive
(since it is negative) of 0.0003 to the current change in the USD Index (reciprocal) values causes a decrease of about
Bangkok SET50 Index. Its effect is different for the recent eight 21724.50 units in the change in the Bangkok SET50 Index.
years for the previous change in the Bangkok SET50 Index
caused an increase of 0.002 to the current change. However, Results for Indonesia
these estimated effects were found to be not statistically
significant. Another non-significant predictor in the model is Association of the Jakarta Stock Exchange Composite Index to
the previous change in THB/USD exchange rate. For both year IDR/USD Exchange Rate and USD Index (reciprocal)
groups, its effect is positive.
Significant effect was determined in the current change in the Table 15. Correlation coefficients and P-values of Jakarta
Stock Exchange to IDRUSD and US Dollar Index
THB/USD exchange rate. A unit increase in the current change
in the THB/USD exchange rate caused an increase of 16455.6 to Association of Pearson’s
the current change in the Bangkok SET50 Index in 2017–2018 Year Stock Market Correlation
and a higher increase of 32884.91 in 2011–2018. This can provide Group to: Coefficient Interpretation p-value
evidence that the Bangkok SET50 Index has linear dependence IDR/USD
0.3312 weak & direct <.001 **
to THB/USD exchange rate. 2017– Exchange Rate
Only around 2.59% of changes in the Bangkok SET50 Index 2018 USD Index very weak & 0.020
can be explained by the three predictors during 2017–2018. 0.1060
(reciprocal) direct **
Extending it to the most recent eight years, the predictors can
IDR/USD
explain 5.78% of variations in the Bangkok SET50 Index. 0.3491 weak & direct <.001 **
2009– Exchange Rate
2018 USD Index very weak &
0.0321 0.114 ns
(reciprocal) direct

PAGE 44 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

The Jakarta Stock Exchange Composite Index (JSX) had a Dependence of Jakarta Stock Exchange Composite Index:
weak (r = 0.3312) direct linear association to IDR/USD Exchange Stationary Tests
rate for the years 2017 to 2018 but shared stronger (r = 0.3491)
direct linear association for the past 10 years. Both were found Table 17. Results of Dickey-Fuller Test for Jakarta Stock
Exchange to IDRUSD and USD Index
to be statistically significant. Its association to USD Index
(reciprocal), on the other hand, is very weak but direct and Dickey-Fuller Test
significant for 2017–2018. This implies that the Jakarta Stock p-value
Exchange Composite Index does share linear association with Year At the First-
the reciprocal of the USD Index (reciprocal) in the short run. All Group Variable level Difference Remark
associations were positive or direct, which means that when First-
IDR/USD exchange rate or USD index (reciprocal) increases (or Jakarta Stock
differenced
decreases), the Jakarta Stock Exchange Composite Index also Exchange 0.667 ns <.001 **
data is
increases (or decreases). Composite Index
stationary.
First-
Dependence of Jakarta Stock Exchange Composite Index: 2017– IDR/USD differenced
Selection of Lag Order 0.709 ns
<.001 **
2018 Exchange Rate data is
stationary.
Table 16. Selected lag orders and LR Test p-value for
Jakarta Stock Exchange to IDRUSD and US Dollar Index First-
USD Index differenced
Selected 0.926 ns
<.001 **
(reciprocal) data is
Year Lag LR Test stationary.
Group Variable Order Suggested by: p-value
First-
Jakarta Stock Jakarta Stock
differenced
Exchange 1 all four <.001 ** Exchange 0.961 ns <.001 **
data is
Composite Index Composite Index
stationary.
2017–
IDR/USD First-
2018 2 all four 0.002 **
Exchange Rate 2009– IDR/USD differenced
0.616 ns
<.001 **
USD Index 2018 Exchange Rate data is
1 all four <.001 ** stationary.
(reciprocal)
Jakarta Stock First-
Exchange 4 all four <.001 ** USD Index differenced
0.484 ns <.001 **
Composite Index (reciprocal) data is
2009– stationary.
IDR/USD
2018 1 HQIC, SBIC <.001 **
Exchange Rate All variables across different timeframes that were tested
USD Index showed that their respective time series is a stationary process.
1 all four <.001 **
(reciprocal) This essentially means that the change in the Jakarta Stock
Exchange Composite Index is somewhat predictable by the
For the most recent two years, the Jakarta Stock Exchange movements in the IDR/USD exchange rate and the USD Index
Composite Index and the USD Index (reciprocal) are somehow (reciprocal) and is not due to “random walk.”
dictated by the most recent previous value (lag of 1), while the
IDR/USD exchange rate is shown to be affected by the two
previous values. For the last eight years, both the IDR/UDS
and the USD Index (reciprocal) are dictated by the most recent
previous value, while the Jakarta Stock Exchange Composite
Index is shown to be affected generally by the four previous
values.

IFTA.ORG PAGE 45
 IFTA JOURNAL 2021 EDITION

Table 19. Regression result for Indonesia (IDRUSD and

Dependence of Jakarta Stock Exchange: Regression Models of USD Index reciprocal versus JSX)
Jakarta Stock Exchange
2017–2018 2009–2018
Table 18. Regression result for Indonesia (IDRUSD versus Predictor Coefficient p-value Coefficient p-value
JSX)
(L1D). Jakarta
2017–2018 2009–2018 -0.04 0.515 ns 0.01 0.732 ns
Composite Index
Predictor Coefficient p-value Coefficient p-value (L2D). Jakarta
- - -0.02 0.500 ns
(L1D). Jakarta Composite Index
-0.03 0.615 ns
0.01 0.747 ns
Composite Index (L3D). Jakarta
-0.10 <.001 **
(L2D). Jakarta Composite Index
- - -0.02 0.508 ns
Composite Index (L4D). Jakarta
-0.04 0.074 **
(L3D). Jakarta Composite Index
- - -0.10 <.001 **
Composite Index (D). IDR/USD
7.83 x 107 <.001 ** 4.40 x 107 <.001 **
(L4D). Jakarta Exchange Rate
- - -0.04 0.072 **
Composite Index (L1D). IDR/USD
-1.41 x 107 0.307 ns 3857481.00 0.214 ns
(D). IDR/USD <.001 Exchange Rate
7.96 x 107 4.39 x 107 <.001 **
Exchange Rate ** (L2D). IDR/USD
1.53 x 107 0.199 ns - -
(L1D). IDR/USD Exchange Rate
-6570148.00 0.634 ns
3724836 0.212 ns
Exchange Rate (D). USD Index
95492.15 0.108 ns -5646.01 0.762 ns
(L2D). IDR/USD (reciprocal)
1.81 x 107 0.111 ns -
Exchange Rate (L1D). USD Index
79895.24 0.174 ns 3053.63 0.831 ns
R-squared 11.71% 13.52% (reciprocal)
R-squared 12.62% 13.52%
Significant effect was determined in the current change
in the IDR/USD exchange rate. A unit increase in the current Incorporating the USD Index into the model showed that
change in the IDR/USD exchange rate caused an increase of significant effect remained to be determined in the current
7.96x107 to the current change in the Jakarta Stock Exchange change in the IDR/USD exchange rate. A unit increase in current
Composite Index in 2017–2018 and an increase of 4.39 x 107 in change in the IDR/USD exchange rate caused an increase of
2009–2018. This can provide evidence that the Jakarta Stock 7.83 x 107 to current change in the Jakarta Stock Exchange
Exchange Composite Index has linear dependence to the IDR/ Composite Index in 2017–2018 and an increase of 4.40 x 107 in
USD exchange rate. 2009–2018. This can likewise provide evidence that the Jakarta
Only around 11.71% of changes in the Jakarta Stock Exchange Stock Exchange Composite Index has linear dependence to the
Composite Index can be explained by the three predictors IDR/USD exchange rate.
during 2017–2018. Extending it to the most recent 10 years, the Only around 12.62% of changes in the Jakarta Stock Exchange
predictors can explain 13.52% of variations in the Jakarta Stock Composite Index can be explained by the above predictors
Exchange Composite Index. during 2017–2018. Extending it to the most recent 10 years, the
predictors can explain 13.52% of variations in the Jakarta Stock
Exchange Composite Index.

PAGE 46 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Results for Malaysia For the most recent two years, the Bursa Malaysia KLCI
Index and the USD Index (reciprocal) are somehow dictated
Association of Bursa Malaysia KLCI Index to MYR/USD by the most recent previous value (lag of 1), while the MYR/
Exchange Rate and USD Index (reciprocal) USD exchange rate is shown to be affected by the two previous
values. For the last 10 years, both the Bursa Malaysia KLCI Index
Table 20. Correlation coefficients and p-values for and MYR/USD exchange rate are dictated by the two most
Malaysia
recent values, while the USD Index (reciprocal) is shown to be
Association of Pearson’s affected by the most recent previous value.
Year Stock Market Correlation
Group to: Coefficient Interpretation p-value Dependence of Bursa Malaysia KLCI Index: Stationary Tests
MYR/USD
0.2349 weak & direct <.001 ** Table 22. Result of Dicker-Fuller Test on Malaysia
2017– Exchange Rate
2018 USD Index very weak & Dickey-Fuller Test
0.0213 0.639 ns p-value
(reciprocal) direct
MYR/USD Year At the First-
0.3663 weak & direct <.001 ** Group Variable level Difference Remark
2009– Exchange Rate
2018 USD Index very weak & First-
0.0213 0.291 ns Bursa Malaysia differenced
(reciprocal) direct 0.417 ns
<.001 **
KLCI Index data is
The Bursa Malaysia KLCI Index had a weak (r = 0.2349) stationary.
direct linear association to the MYR/USD Exchange rate for
First-
the years 2017 to 2018 but shared stronger (r = 0.3663) direct
2017– MYR/USD differenced
linear association for the past 10 years. Both were found to be 0.989 ns
<.001 **
2018 Exchange Rate data is
statistically significant. On the other hand, its associations stationary.
to the USD Index, aside from being very weak, were also not
First-
significant. This implies that the Bursa Malaysia KLCI Index
USD Index differenced
does not share linear association with USD Index (reciprocal). 0.926 ns <.001 **
(reciprocal) data is
All associations were positive or direct, which means that when stationary.
the MYR/USD exchange rate or USD index (reciprocal) increases
First-
(or decreases), the Bursa Malaysia KLCI Index also increases (or
Bursa Malaysia differenced
decreases). KLCI Index
0.993 ns <.001 **
data is
stationary.
Dependence of Bursa Malaysia KLCI Index: Selection of Lag
Order First-
2009– MYR/USD differenced
0.632 ns <.001 **
Table 21. Selected lag order and LR Test p-value for 2018 Exchange Rate data is
Malaysia stationary.
First-
Year Selected LR Test
Group Variable Lag Order Suggested by: p-value USD Index differenced
0.490 ns
<.001 **
(reciprocal) data is
Bursa Malaysia stationary.
1 HQIC, SBIC <.001 **
KLCI Index
All variables across different timeframes that were tested
2017– MYR/USD
2 all four <.001 ** showed that their respective time series is a stationary process.
2018 Exchange Rate
This essentially means that the change in Bursa Malaysia KLCI
USD Index Index is somewhat predictable by the movements in the MYR/
1 all four <.001 **
(reciprocal)
USD exchange rate and is not due to “random walk.”
Bursa Malaysia
2 all four <.001 **
KLCI Index
2009– MYR/USD
2 HQIC, SBIC <.001 **
2018 Exchange Rate
USD Index
1 all four <.001 **
(reciprocal)

IFTA.ORG PAGE 47
 IFTA JOURNAL 2021 EDITION

Dependence of Bursa Malaysia KLCI Index: Regression Models Incorporating the USD Index (reciprocal) into the model
of Bursa Malaysia KLCI Index showed that significant effect remained to be determined in the
current change in MYR/USD exchange rate. A unit increase in
Table 23. Regression result for Malaysia (MYRUSD versus current change in MYR/USD exchange rate caused an increase
Bursa Malaysia)
of 4778.66 to the current change in the Bursa Malaysia KLCI
2017–2018 2009–2018 Index in 2017-2018 and an increase of 2655.89 in 2009-2018. This
Predictor Coefficient p-value Coefficient p-value
can likewise provide evidence that Bursa Malaysia KLCI Index
has linear dependence to MYR/USD exchange rate.
(L1D). Bursa Only around 6.13% of changes in Bursa Malaysia KLCI Index
Malaysia KLCI 0.06 0.408 ns 0.06 0.050 ** can be explained by the above predictors during 2017–2018.
Index
Extending it to the most recent 10 years, the predictors can
(L2D). Bursa explain 14.14% of the variations in the Bursa Malaysia KLCI
Malaysia KLCI - - -0.02 0.544 ns Index.
Index
(D). MYR/USD Results for India
4639.47 <.001 ** 2577.79 <.001 **
Exchange Rate
(L1D). MYR/USD Association of the BSE SENSEX to INR/USD Exchange Rate and
-357.71 0.696 ns -97.75 0.548 ns USD Index (reciprocal)
Exchange Rate
(L2D). MYR/USD Table 25. Correlation coefficients and p-values for India
-45.97 0.960 ns 336.32 0.031 **
Exchange Rate
Association of Pearson’s
R-squared 5.82% 13.86% Year Stock Market Correlation
Group to: Coefficient Interpretation p-value
Significant effect was determined in the current change in INR/USD
0.3489 weak & direct <.001 **
MYR/USD exchange rate. A unit increase in current change 2017– Exchange Rate
in MYR/USD exchange rate caused an increase of 4639.47 to 2018 USD Index very weak &
current change in the Bursa Malaysia KLCI Index in 2017–2018 0.0675 0.134 ns
(reciprocal) direct
and an increase of 2577.79 in 2009–2018. This can provide
evidence that Bursa Malaysia KLCI Index has linear dependence INR/USD
0.3832 weak & direct <.001 **
2011– Exchange Rate
to MYR/USD exchange rate.
2018 USD Index very weak &
0.0004 0.985 ns
Around 5.82% only of changes in Bursa Malaysia KLCI Index (reciprocal) direct
can be explained by the three predictors during 2017–2018.
Extending it to the most recent 10 years, the predictors can The BSE SENSEX had a weak (r = 0.3489) direct linear
explain 13.86% of variations in the Bursa Malaysia KLCI Index. association to INR/USD Exchange rate for the years 2017 to 2018
but shared stronger (r = 0.3832) direct linear association for the
Table 24. Regression result for Malaysia (MYRUSD and past 10 years. Both were found to be statistically significant. On
USD Index reciprocal versus Bursa Malaysia)
the other hand, its association to USD Index (reciprocal) in both
2017–2018 2009–2018 timeframes are both very weak and not significant. This implies
Predictor Coefficient p-value Coefficient p-value
that the BSE SENSEX does not share a linear association with the
USD Index (reciprocal). All associations were positive or direct,
(L1D). Bursa which means that when the INR/USD exchange rate increases
0.06 0.378 ns 0.06 0.049 **
Malaysia KLCI Index (or decreases), the BSE SENSEX also increases (or decreases).
(L2D). Bursa
- -0.02 0.551 ns
Malaysia KLCI Index
(D). MYR/USD
4778.66 <.001 ** 2655.89 <.001 **
Exchange Rate
(L1D). MYR/USD
-241.48 0.814 ns 29.00 0.866 ns
Exchange Rate
(L2D). MYR/USD
-472.05 0.615 ns 334.11 0.040 **
Exchange Rate
(D). USD Index
-8183.29 0.480 ns -9172.82 0.006 **
(reciprocal)
(L1D). USD Index
11455.66 0.337 ns -685.82 0.832 ns
(reciprocal)
R-squared 6.13% 14.14%

PAGE 48 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Dependence of BSE SENSEX: Selection of Lag Order Dependence of BSE SENSEX: Regression Models of BSE SENSEX
Table 26. Selected lag order and LR Test p-value for India Table 28. Regression result for India (INRUSD versus BSE
SENSEX)
Year Selected LR Test
Group Variable Lag Order Suggested by: p-value 2017-2018 2011-2018
BSE SENSEX 1 all four <.001 ** Predictor Coefficient p-value Coefficient p-value
INR/USD (L1D). BSE SENSEX 0.04 0.497 ns
0.05 0.055 **
2017– 1 all four <.001 **
Exchange Rate
2018 (L2D). BSE SENSEX - - -0.03 0.249 ns
USD Index
1 all four <.001 ** (D). INR/USD
(reciprocal) 1573548.00 <.001 ** 1055073.00 <.001 **
Exchange Rate
BSE SENSEX 2 all four 0.001 **
(L1D). INR/USD
-9195.62 0.965 ns -8908.05 0.892 ns
INR/USD Exchange Rate
2011– 3 FPE, AIC, HQIC 0.001 **
Exchange Rate
2018 (L2D). INR/USD
- - 70482.62 0.264 ns
USD Index 0.020 Exchange Rate
2 FPE, AIC, HQIC
(reciprocal) **
(L3D). INR/USD
- - 132600.10 0.024 **
For the most recent two years, the BSE SENSEX, INRUSD Exchange Rate
exchange rate, and USD Index (reciprocal) are somehow dictated R-squared 12.33% 15.38%
by the most recent previous value (lag of 1). For the last eight
years, both the BSE SENSEX and the USD Index are dictated by Significant effect was determined in the current change in
the two most recent values, while the INR/USD exchange rate is INR/USD exchange rate. A unit increase in current change in
shown to be affected by the three most recent values. the INR/USD exchange rate caused an increase of 1573548.00
to current change in the BSE SENSEX in 2017-2018 and an
Dependence of BSE SENSEX: Stationary Tests increase of 1055073.00 in 2009-2018. This can provide
evidence that the BSE SENSEX has linear dependence to the
Table 27. Result of Dicker-Fuller Test on India TRY/USD exchange rate.
Dickey-Fuller Test Only around 12.33% of changes in BSE SENSEX can be
p-value explained by the three predictors during 2017–2018. Extending
Year At the First- it to the most recent 10 years, the predictors can explain 15.38%
Group Variable level Difference Remark of variations in the BSE SENSEX.
First-differenced Table 29. Regression result for India (INRUSD and USD
BSE SENSEX 0.974 ns <.001 **
data is stationary. Index reciprocal versus BSE SENSEX)
INR/USD 2017–2018 2011–2018
2017– First-differenced
Exchange 0.726 ns <.001 **
2018 data is stationary. Predictor Coefficient p-value Coefficient p-value
Rate
USD Index First-differenced (L1D). BSE SENSEX 0.04 0.492 ns
0.05 0.056 **
0.944 ns <.001 **
(reciprocal) data is stationary. (L2D). BSE SENSEX - - -0.03 0.218 ns
First-differenced (D). INR/USD
BSE SENSEX 0.582 ns <.001 ** 1599073.00 <.001 ** 1120737.00 <.001 **
data is stationary. Exchange Rate
INR/USD (L1D). INR/USD
2011– First-differenced 34645.80 0.872 ns 26361.22 0.711 ns
Exchange 0.987 ns <.001 ** Exchange Rate
2018 data is stationary.
Rate
(L2D). INR/USD
USD Index First-differenced - - 84261.07 0.208 ns
0.811 ns <.001 ** Exchange Rate
(reciprocal) data is stationary.
(L3D). INR/USD
- - 137791.70 0.020 **
All variables across different timeframes that were tested Exchange Rate
showed that their respective time series is a stationary process. (D). USD Index
This essentially means that the change in the BSE SENSEX -125281.40 0.602 ns -445374.10 <.001 **
(reciprocal)
is somewhat predictable by the movements in the INR/USD
(L1D). USD Index
exchange rate and is not due to “random walk.” -168917.50 0.507 ns -94560.75 0.364 ns
(reciprocal)
(L2D). USD Index
- - -57774.91 0.541 ns
(reciprocal)
R-squared 12.44% 16.42%

IFTA.ORG PAGE 49
 IFTA JOURNAL 2021 EDITION

Incorporating the USD Index (reciprocal) into the model Dependence of Borsa Istanbul 100 Index: Selection of Lag Order
showed that significant effect remained to be determined in
the current change in INR/USD exchange rate. A unit increase Table 31. Selected lag order and LR Test p-value for
Turkey
in the current change in the INR/USD exchange rate caused an
increase of 1599073.00 to the current change in the BSE SENSEX Selected
in 2017–2018 and an increase of 1120737.00 in 2011–2018. Year Lag LR Test
This can likewise provide evidence that BSE SENSEX has linear Group Variable Order Suggested by: p-value
dependence to the INR/USD exchange rate. Borsa Istanbul
1 all four <.001 **
Only around 12.44% of changes in BSE SENSEX can be 100 Index
explained by the above predictors during 2017–2018. Extending 2017– TRY/USD 0.020
it to the most recent eight years, the predictors can explain 4 FPE, AIC, HQIC
2018 Exchange Rate **
16.42% of variations in the BSE SENSEX.
USD Index
1 all four <.001 **
(reciprocal)
Results for Turkey
Borsa Istanbul
1 all four <.001 **
Association of Borsa Istanbul 100 Index to TRY/USD Exchange 100 Index
Rate and USD Index (reciprocal) 2009– TRY/USD
4 all four 0.001 **
2018 Exchange Rate
Table 30. Correlation coefficients and p-values for
Turkey USD Index
1 all four <.001 **
(reciprocal)
Association of Pearson’s
Year Stock Market Correlation For the most recent two years, the Borsa Istanbul 100 Index
Group to: Coefficient Interpretation p-value and the USD Index (reciprocal) are somehow dictated by
TRY/USD the most recent previous value (lag of 1), while the TRY/USD
0.3556 weak & direct <.001 **
2017– Exchange Rate exchange rate is shown to be affected by the four previous
2018 USD Index very weak & values. For the last 10 years, both the Borsa Istanbul 100 Index
0.1675 <.001 ** and the USD Index are dictated by the most recent value, while
(reciprocal) direct
the TRY/USD exchange rate is shown to be affected by the four
TRY/USD
0.3624 weak & direct <.001 ** most recent values.
2009– Exchange Rate
2018 USD Index very weak &
0.1192 <.001 ** Dependence of Borsa Istanbul 100 Index: Stationary Tests
(reciprocal) direct
Table 32. Result of Dicker-Fuller Test on Turkey
The Borsa Istanbul 100 Index had a weak (r = 0.3556) direct Dickey-Fuller Test
linear association to the TRY/USD Exchange rate for the Year p-value
years 2017 to 2018 but shared stronger (r = 0.3624) direct Variable Remark
Group At the First-
linear association for the past 10 years. Both were found to be level Difference
statistically significant. On the other hand, its association to
Borsa
the USD Index (reciprocal) in both timeframes is very weak but Istanbul 100 0.877 ns <.001 **
First-differenced
nonetheless significant. This implies that the Borsa Istanbul 100 data is stationary.
Index
Index does share to some degree a linear association with the
USD Index (reciprocal). All associations were positive or direct, 2017– TRY/USD
First-differenced
2018 Exchange 0.652 ns <.001 **
which means that when the TRY/USD exchange rate or the USD data is stationary.
Rate
index (reciprocal) increases (or decreases), the Borsa Istanbul
100 Index also increases (or decreases). USD Index First-differenced
0.939 ns <.001 **
(reciprocal) data is stationary.
Borsa
First-differenced
Istanbul 100 0.903 ns <.001 **
data is stationary.
Index
2009– TRY/USD
First-differenced
2018 Exchange 0.601 ns <.001 **
data is stationary.
Rate
USD Index First-differenced
0.488 ns <.001 **
(reciprocal) data is stationary.

PAGE 50 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Table 34. Regression result for Turkey (TRYUSD and USD

All variables across different timeframes that were tested Index reciprocal versus Borsa Istanbul)
showed that their respective time series is a stationary process.
2017–2018 2009–2018
This essentially means that the change in Borsa Istanbul 100
Index is somewhat predictable by the movements in the TRY/ Predictor Coefficient p-value Coefficient p-value
USD exchange rate and is not due to “random walk.” (L1D). Borsa
-0.004 0.946 ns -0.08 0.005 **
Istanbul 100 Index
Dependence of Borsa Istanbul 100 Index: Regression Models of
(D). TRY/USD
Borsa Istanbul 100 Index Exchange Rate
148116.60 <.001 ** 108792.00 <.001 **

Table 33. Regression result for Turkey (TRYUSD versus (L1D). TRY/USD
Borsa Istanbul) -20644.26 0.468 ns 19100.34 0.006 **
Exchange Rate
2017–2018 2009–2018 (L2D). TRY/USD
4007.38 0.844 ns -4058.99 0.422 ns
Exchange Rate
Predictor Coefficient p-value Coefficient p-value
(L3D). TRY/USD
(L1D). Borsa -23985.19 0.169 ns -805.70 0.865 ns
0.01 0.903 ns -0.08 0.004 ** Exchange Rate
Istanbul 100 Index
(L4D). TRY/USD
(D). TRY/USD 11836.40 0.507 ns -5431.19 0.270 ns
157071.60 <.001 ** 101407.30 <.001 ** Exchange Rate
Exchange Rate
(D). USD Index
(L1D). TRY/USD 1969789.00 0.148 ns -1053689.00 0.010 **
-20281.53 0.450 ns 17330.87 0.003 ** (reciprocal)
Exchange Rate
(L1D). USD Index
(L2D). TRY/USD 897357.40 0.515 ns -272519.20 0.472 ns
3030.97 0.882 ns -4339.60 0.391 ns (reciprocal)
Exchange Rate
R-squared 13.60% 14.18%
(L3D). TRY/USD
-24735.01 0.156 ns -1031.25 0.827 ns
Exchange Rate
Incorporating the USD Index (reciprocal) into the model
(L4D). TRY/USD showed that significant effect remained to be determined in the
11432.70 0.522 ns -5184.45 0.290 ns
Exchange Rate current change in the TRY/USD exchange rate. A unit increase
R-squared 13.17% 13.90% in the current change in TRY/USD exchange rate caused an
increase of 148116.60 to the current change in the Borsa
Significant effect was determined in the current change in Istanbul 100 Index in 2017–2018 and an increase of 108792.00
the TRY/USD exchange rate. A unit increase in current change in 2009–2018. This can likewise provide evidence that the
in TRY/USD exchange rate caused an increase of 157071.60 Borsa Istanbul 100 Index has linear dependence to the TRY/USD
to the current change in the Borsa Istanbul 100 Index in exchange rate.
2017–2018 and an increase of 101407.30 in 2009–2018. This can Only around 13.60% of changes in the Borsa Istanbul 100
provide evidence that the Borsa Istanbul 100 Index has linear Index can be explained by the above predictors during 2017–
dependence to the TRY/USD exchange rate. 2018. Extending it to the most recent 10 years, the predictors
Only around 13.17% of changes in the Borsa Istanbul 100 Index can explain 14.18% of variations in the Borsa Istanbul 100 Index.
can be explained by the three predictors during 2017–2018.
Extending it to the most recent 10 years, the predictors can Results for MSCI Emerging Markets
explain 13.90% of variations in the Borsa Istanbul 100 Index.
Correlation of MSCI Emerging Markets Index to MSCI EM
Currency Index
Table 35. Correlation coefficients and p-values for MSCI
EM Index vs. MSCI EM Currency Index

Pearson’s
Correlation
Year Group Coefficient Interpretation p-value
2017–2018 0.9418 very strong & direct <.001 **
2015–2019 0.9673 very strong & direct <.001 **

The MSCI Emerging Markets Index had a very strong (r =

0.9418) direct linear association to the MSCI EM Currency Index
rate for the years 2017 to 2018 and shared an even stronger (r
= 0.9673) direct linear association for the past five years. Both
were found to be statistically significant. This implies that the
MSCI Emerging Markets Index shares to a huge degree a linear
association with the MSCI EM Currency Index. All associations

IFTA.ORG PAGE 51
 IFTA JOURNAL 2021 EDITION

were positive or direct, which means that when MSCI EM Tally of “Signals” generated from the MSCI EM Currency Index
Currency Index increases (or decreases), the MSCI Emerging
Market Index also increases (or decreases). Figure 1. Chart of MSCI EM Index versus MSCI EM
Currency Index (Jan. 12, 2015, to April 1, 2019)
Dependence of MSCI Emerging Markets Index to MSCI EM
Currency Index
Table 36. Regression result for MSCI EM Index versus
MSCI EM Currency Index

2017–2018 2015–2019
Predictor Coefficient p-value Coefficient p-value
(L1D). MSCI
Emerging Market 0.11 0.011 ** 0.10 0.002 **
Index
Using the daily data from January 12, 2015, to April 1, 2019,
(L2D). MSCI it was found that there were 17 instances where the MSCI EM
Emerging Market 0.05 0.101 ns -0.02 0.542 ns Currency Index declined by 2% from an established swing high.
Index With approximately 20 trading days of duration, the said 2%
(D). MSCI EM or so drop in MSCI EM Currency Index translated to an average
1.34 <.001 ** 1.29 <.001 **
Currency Index 4.64% decline in the EM equity index in the same period. The
(L1D). MSCI EM MSCI EM Index continued to decline by another 1.10% on average
-0.13 0.136 ns -0.08 0.161 ns in the succeeding month. Two months post the 2% dip in the
Currency Index
currency basket, the EM equity index has remained weak,
(L2D). MSCI EM
- - 0.09 0.114 ns dropping by another 1.34% on average. Three months post the
Currency Index
2% dip in the currency basket, the EM equity index was down by
(L3D). MSCI EM an average of 3.67% on top of its initial decline.
- - 0.03 0.452 ns
Currency Index
There were also 14 instances where the MSCI EM Currency
R-squared 46.91% 53.74% Index broke down by 2% from an established swing low. With
approximately 44 trading days in duration, the said 2% or so slip
Significant effect was determined in the current change in the in the EM currency basket translated to an average 6.69% decline
MSCI EM Currency Index. A unit increase in the current change in the EM equity index in the same period. The EM equity index
in the MSCI EM Currency Index caused an increase of 1.34 to the continued to falter by another 3.22% on average in the succeeding
current change in the MSCI Emerging Markets Index in 2017– month. Two months post the 2% dip from the swing low of the
2018 and an increase of 1.29 in 2015–2019. This can provide currency basket, the EM equity index has mildly recovered but
notable evidence that MSCI Emerging Markets Index has linear remained down by another 1.53% on average. Three months post
dependence to the MSCI EM Currency Index. the 2% dip in the currency basket, the EM equity index rallied by
2.92% on average from its low on the 44th day.
Around 46.91% of the changes in the MSCI Emerging Markets In the 31 instances where the MSCI EM Currency Index
Index can be explained by the two predictors during 2017–2018. declined by 2% or so from both swing high and swing low with
Extending it to the most recent five years, the predictors can an average duration of 30 trading days, the MSCI Emerging
explain 53.74% of variations in the MSCI Emerging Markets Markets Index consequently declined by an average of 5.22% in
Index. the same period. The EM equity index continued to weaken by
another 2.51% on average one month after. In the two months
after the 2% dip in the currency basket, the EM equity index
recovered a bit but remained down by another 1.78%. In the
three months after the 2% dip in the currency basket, the EM
equity index continued to recover but nonetheless remained
down by 1.02% on average on top of its initial slip.

Table 37. Percentage changes following swing high declines and swing low breaks (MSCI)

% Decline in EM Equities
2% Slip in EM Currency Index # Signals Duration in Days During 1 Month After 2 Months After 3 Months After
Swing High Decline 11 19.82 -4.64% -1.10% -1.34% -3.67%
Swing Low Break 8 44.25 -6.69% -3.22% -1.53% 2.92%
Total 19 30.11 -5.22% -2.51% -1.78% -1.02%

PAGE 52 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Figure 2. Graph of % changes in EM equities following

swing high declines and swing low breaks explain as much as 53% of the changes in the latter. Even if the
changes in the currency only explains 12–14% of the changes in
its respective equities covering 8 to 10 years, the author deems
that such is already heavy enough to warrant a closer analysis
of the movements in the exchange rate when actively managing
a domestic equity portfolio. Adverse domestic currency
movements can indeed be used as an indicator of a potential
earthquake in the corresponding equity space both in the one to
three months ahead and in the long run.

References
Adam, Shamim and others, The Story of Malaysia’s 1MDB, The Scandal
That Shook the World of Finance, Bloomberg, Bloomberg L.P., 24 May
2018, bloomberg.com/news/articles/2018-05-24/how-malaysia-s-
1mdb-scandal-shook-the-financial-world-quicktake
Discussion Aghevlie, Bijan B., The Asian Crisis: Causes and Remedies, IMF,
This paper provided an argument and ample evidence to International Monetary Fund, June 1999, imf.org/external/pubs/ft/
support the use of the changes in the Asian emerging market fandd/1999/06/aghevli.htm
currencies as signals for forecasting consequent changes in Baccardax, Martin, Turkey’s Currency Crisis: What Happened, What’s
their respective equity markets. While from the correlation Next, Why It Matters, TheStreet, The Street, Inc., 10 August 2018,
analysis that was used covering the periods of the last one year thestreet.com/markets/turkey-s-currency-crisis-what-happened-
and the last 8 to 10 years showed a weak linear relationship what-s-next-why-it-matters-14679835
between the domestic Asian currencies against their respective Carson, Michael, and John Clark, Asian Financial Crisis, Federal Reserve
equity indices, their association nonetheless remained History, Federal Reserve Bank of Richmond, 22 November 2013,
statistically significant. The analysis and the results of the federalreservehistory.org/essays/asian_financial_crisis
“signals” that were generated actually supports the findings of Chin, Helen, Chinese Yuan Depreciates in August as a Result of
Usman Umer et al. in that the “comovements between exchange Exchange Rate Reform, Fung Business Intelligence, The Fung
rates and stock prices become stronger during the crises time” Intelligence Centre, 19 August 2015, fbicgroup.com/sites/default/
or when there was an increased volatility in the exchange files/Chinese%20yuan%20depreciates%20in%20August%2019%20
rate in the short run. It was determined that the sensitivity or AUG%202015.pdf
beta of equities from the movement in the exchange rate can Corsetti, Giancarlo, et al. What Caused the Asian Currency and
be as much as 2.6 times. Furthermore, it was also seen that Financial Crisis?, Federal Reserve Bank of New York, Federal Reserve
the resulting weakness in equities have not just persisted but Bank of New York, January 1998, newyorkfed.org/medialibrary/
even deepened in the succeeding one to three months. One media/research/economists/pesenti/whatjapwor.pdf
may therefore use a “breakdown” in the domestic currency Data and Ticker Symbols, Investing.com, Web. April 2019
as a signal to hedge against a possible drop of as much as 10% Data and Ticker Symbols, TradingEconomics.com, Web. April 2019
in the respective domestic equities in the next two and half
months. Additionally, it was determined that the association Dornbusch, Rudiger, and Stanley Fischer, Exchange Rates and the
Current Account, American Economic Association, December 1980.
and proportion of the variance (R squared) between domestic
Web. May 2019
currencies and equity markets notably improve with longer
historical data of 8 to 10 years relative to a year’s data alone. Dumlao Jr., Santiago F., The Asian Financial Crisis (AFC), Part One.
Thailand, BusinessMirror, The BusinessMirror, 6 June 2017,
businessmirror.com.ph/2017/06/06/the-asian-financial-crisis-afc-
Conclusion part-one-thailand/
For the periods covering the last 8 to 10 years, the stationary
tests that were done all showed that the sample Asian equity —, The Asian Financial Crisis (AFC), Part Two. The Philippine Response,
indices were to a certain degree predictable by the movements BusinessMirror, The BusinessMirror, 4 July 2017, businessmirror.
com.ph/2017/07/04/the-asian-financial-crisis-part-two-the-
in their domestic currency and, more importantly, that their
philippine-response/
movements were not due to “random walk.” It was also found
that significant effect was determined in the current change in Farrer, Martin, China Devalues Yuan by 2% to Boost Flagging Economy,
the domestic currencies vis a vis their respective equity index The Guardian, Guardian News & Media Limited, 11 August 2015,
theguardian.com/business/2015/aug/11/china-devalues-yuan-by-2-
using regression analysis. While the changes in the currency
to-boost-flagging-economy
only explain as low as 2.58% in the past year that was tested and
5.78% in the past eight years the changes in the equity index General Information on the MSCI Emerging Markets Index, MSCI, MSCI,
as in the case of Thailand, the dependence can be found much Inc., msci.com/emerging-markets
higher in Indonesia, Malaysia, India, and Turkey. Meanwhile, General Information on the MSCI Emerging Markets Currency Index,
the correlation and regression analyses done in the MSCI EM MSCI, MSCI, Inc., msci.com/documents/10199/0b995d45-15ef-4570-
Currency Index against the MSCI EM Index actually found a very 9aa0-ef28baa89083
strong direct correlation of at least 94%, while the former could

IFTA.ORG PAGE 53
 IFTA JOURNAL 2021 EDITION

Granger, Clive, and others, A Bivariate Causality between Stock Prices —, Trading With Intermarket Analysis: A Visual Approach to Beating the
and Exchange Rates: Evidence from Recent Asia Flu, University of Financial Markets Using Exchange-Traded Funds, John Wiley & Sons,
California, San Diego, 1 April 1998. Web. May 2019 Inc., 2012
Harper, Jo, Turkish Lira Hits Record Low as US Sanctions Bite, DW, Qayyum, Abdul, and A.R. Kemal, Volatility Spillover between the Stock
Deutsche Welle, 3 August 2018, dw.com/en/turkish-lira-hits-record- Market and the Foreign Market in Pakistan, Pakistan Institute of
low-as-us-sanctions-bite/a-44943764 Development Economics, 2006. Web. May 2019
Holland, Tom, Turkey’s Crisis Shows a Deeper Disease – and Asia is Riley, Charles, Yuan Tumbles for Second Day as China Devalues, CNN
Infected Too, South China Morning Post, South China Morning Post International, Cable News Network, 12 August 2015, money.cnn.
Publishers Ltd., 20 August 2018, scmp.com/week-asia/opinion/ com/2015/08/11/investing/yuan-china-rmb-currency/index.html
article/2160222/turkeys-crisis-shows-deeper-disease-and-asia- Sengupta, Ahona, As Rupee Slumps to Record Low of 71, Here’s Why
infected-too It’s Falling Like Never Before, News 18, Network18 Media and
Inman, Phillip, China Plays Down Devaluation Fears as Yuan Cut for Investments Ltd., 31 August 2018, news18.com/news/business/
Third Straight Day, The Guardian, Guardian News & Media Limited, heres-why-the-rupee-is-falling-like-never-before-1862761.html
13 August 2015, theguardian.com/business/2015/aug/13/china- Singh, Hemant, Why Rupee is Falling Against Dollar Currently?, Jagran
plays-down-devaluation-fears-as-yuan-cut-for-third-straight-day Josh, Jagran Prakashan Ltd., 2 April 2019, jagranjosh.com/general-
Lee, Yen Nee, What Went Wrong for Turkey? Its Economy is ‘In the knowledge/why-rupee-is-falling-against-dollar-1530858596-1
Midst of a Perfect Storm’, CNBC, CNBC LLC., 13 August 2018, cnbc. Umer, Usman, and others, The Dynamic Linkages between Exchange
com/2018/08/13/turkey-crisis-economy-faces-weak-lira-inflation- Rates and Stock Prices: Evidence from Emerging Markets, Journal of
debt-and-tariffs.html Finance and Investment Analysis, 2015. Web. May 2019
Ma, Christopher, and G. Wenchi Kao, On Exchange Rate Changes and Vaswani, Karishma, Why China is Devaluing the Yuan, BBC, BBC, 11
Stock Price Reactions, Journal of Business Finance & Accounting, August 2015, bbc.com/news/business-33858447
1990. Web. May 2019
Wright, Tom, Fund Controversy Threatens Malaysia’s Leaders, The Wall
Mishra, Alok, and others, Volatility Spillover between Stock and Foreign Street Journal, Dow Jones & Company, Inc., 18 June 2015, wsj.com/
Exchange Markets: Indian Evidence, International Journal of articles/fund-controversy-threatens-malaysias-leader-1434681241
Business, 2007. Web. May 2019
Wright, Tom, and Simon Clark, Investigators Believe Money Flowed
Mo, Jl, Chinese Yuan: A Shockingly Sharp Depreciation, Amundi Asset to Malaysian Leader Najib’s Accounts Amid 1MDB Probe, The Wall
Management, Amundi, 14 September, 2015, research-center.amundi. Street Journal, Dow Jones & Company, Inc., 2 July 2015, wsj.com/
com/page/Article/2015/09/Chinese-Yuan-a-shockingly-sharp- articles/SB10130211234592774869404581083700187014570
depreciation#Download
Y-Sing, Liau, Ringgit Falls to Five-Year Low as Oil Slump Cuts Revenue
Murphy, John, Intermarket Analysis: Profiting from Global Market Outlook, Bloomberg, Bloomberg L.P., 3 December 2014, bloomberg.
Relationships, John Wiley & Sons, Inc., 2004 com/news/articles/2014-12-03/ringgit-falls-to-five-year-low-as-oil-
—, Intermarket Technical Analysis: Trading Strategies for the Global slump-cuts-revenue-outlook
Stock, Bond, Commodity, and Currency Markets, John Wiley & Sons,
Inc., 1991

PAGE 54 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Coefficient Moving Average Mohamed Fawzy ElSayed Ali AbdAlla,

CETA, CFTe, MFTA
By Mohamed Fawzy ElSayed Ali AbdAlla, CETA, CFTe, MFTA +201005004322
[email protected]

Abstract moving average, is the real average of the data over the period
Data averaging is not only one of the most important but as being the arithmetic mean, which is equal to the summation
also the most commonly used tool in all money-related fields; of the data divided by the number of the data.
from upwards to downwards, it is used in econometrics to SMA(t) = Σ(Pi) / n
analyse a country’s macro-economic factors. It is also applicable Where Pi is the price at a time point (i), n is the number of prices
to fundamental and technical analysis for financial ratios or the time points.
and price charts of sectors and companies. Data average is However, SMA has disadvantages:
performed by many different types and techniques; the most 1. It gives equal weights to all the data from the recent
used and popular are the simple arithmetic, the geometric, the significant datum until the old less important datum.
weighted, and the exponential average. Comparing a sequence 2. It is affected by extreme value, as only one extremely larger
of successive averages is known as moving average. or extremely smaller value will drag the average to it and
Before using a popular tool and depending on its outcome, it away from the rest of the repeating values. This disadvantage
is logically mandatory to understand deeply the advantages and occurs twice; first, when the extreme value has recently come
disadvantages of each type of moving average. The majority of into the average window, and second, when it is late to drop off
technicians consider the exponential moving average (EMA) the period of the average. Mainly, when it exists, the average
to be the most suitable and efficient average in terms of its causes a lot of misleading average movmente. This flaw is
concept, which some believe to be a myth. Moreover, the EMA worse with a shorter moving average (Alexander Elder, 2014).
is embedded in the calculation of many popular technical 3. It does not contain the historical data.
indicators and is used by technicians in coding the formula of The weighted moving average (WMA) gives less weight (=1) to
their developed indicators. the oldest datum in the period that increases regularly (by +1)
This paper will attempt to deeply understand the drawbacks until the most important recent datum in the period.
of the EMA and introduce the the coefficient moving average WMA(t) = Σ(Pi.Wi) / n
(CMA)—a moving average with a new concept in its calculation Where Pi is the price at a time point (i), Wi is the weight of the
that is thought to be a better combination, with more real price Pi, n is the number of prices or the time points.
advantages and fewer disadvantages than other well-known Therefore, WMA overcomes the disadvantages of not giving
moving averages. Besides, CMA is more flexible and able to be more importance to the recent data. In addition, the extreme
adjusted by the user to fit its intended use. value will greatly affect the average only when entering it,
which is accepted, but its weight will decrease gradually until
Introduction exiting the average unnoticed. However, the WMA still does not
contain the historical data as with the SMA.
Overview on Moving Averages in Technical The exponential moving average (EMA) not only gives more
Analysis importance to the recent data but also contains the historical
Moving averages seem to be the most successful tool in data.
the objective of technical analysis—mainly because averages EMA(t) = [ K .P(t) ] + [ (1-K) .EMA(t-1) ]
as lagging indicators profit from the trend by allowing entry K = 2 / (n-1)
and exit points after sufficient confirmation and dampen the Where K is the smoothing coefficient of the EMA, P(t) is
effects of short-term oscillations. Additionally, the averages the price at a time of EMA, EMA(t-1) is the previous EMA,
are essential components in the calculation of almost all other n is the period of the average. Huston has described the
indicators and oscillators, e.g., MACD, Stochastic, CCI, ADX, RSI smoothing constant “k” (Perry Kaufman, 2013).
(Kirkpatrick, 2016).
Therefore, it is mandatory to fully understand the drawbacks Weighted and Exponential Moving
of each of the most commonly used moving averages—simple, Averages Explained
weighted, and exponentia—to know how much we were misled
due to their limitations and how restricted we were in our choices. Digging Deeper Into WMA Misses
The WMA is weighted by multiples of the weight of the oldest
Popular Commonly Used Moving Averages (SMA, datum, which is acceptable. The progress of the cumulative
WMA, EMA) weight for the data weights of a WMA with a period of 100,
The simple moving average (SMA), also named truncated starting from the highest weight of the most recent datum (100)

IFTA.ORG PAGE 55
 IFTA JOURNAL 2021 EDITION

and cumulating to the smallest weight of the oldest datum (1), is And since the weights are the arithmetic series of the natural
illustrated in Table 1 and is compared to the linear progressing numbers “1, 2, 3, 4, 5, …. , n” its summation is
cumulative weight of SMA in Figure 1. = (n²+n)/2 = (n * (n+1)) / 2.
Thus, the weight of the current datum = n / [ (n * (n+1)) / 2 ] =
Table 1. The relation between the number of datum and 2n / (n * (n+1) = 2/(n+1), which is equal to the “K” constant used
the percentage completed from the whole weights of
in the EMA calculation.
WMA
The earlier conclusion means that in the EMA technique:
Number of Completed % of the Number of Completed % of 1. The weight of the most recent datum for the WMA and EMA
Datum WMA weights Datum the WMA weights are equal, and both are equal to 2/(n+1) which is the (K).
14 25.9% 43 67.3% 2. In the WMA, the weight of the previous datum is calculated
again by decreasing rate and so on until the oldest datum
19 34.2% 51 75.7%
takes the weight of 2 / (n²+n). While in the EMA, the weight
25 43.6% 67 88.9% of the previous datum is the same as the weight of the first
30 50.8% 75 93.6% datum, but it is calculated from the residual of the average
weights. And so on, until the oldest datum of the period takes
34 56.2% 91 99.1%
the weight from the last residual of the average weights, and
then the left is set for the historical data weight (data from
Figure 1. The progress of the cumulative weight for a the start of the chart until before the calculated period).
WMA 100 compared to that of the equal weights SMA 100
3. Accordingly, the weight of the historical data takes nothing
from the weight of the most recent datum in the EMA if
compared to the weight of the current datum in the similar
WMA.

Question 2: What is the weight of the historical data (before the

period) in the average weight?
By understanding the calculation of the EMA; where the
recent datum takes (K) weight, and the previous average takes
(1-K) weight. From the weight of the previous average, which
is (1-K), again the before previous datum takes (K) weight and
Therefore, it is evident, for example, that half the data the before previous average takes (1-K) weight. And so on, this
contributes to three-quarters the weight of the average and the process is repeated until the oldest datum in the period takes
oldest tenth of data is nearly neglected. Besides, it is worthy (K) weight, and the last residual (1-K) weight is left for the data
to note that the data in the previous table and chart also before the average period, which stands for all the previous
applies for WMA with periods of 20 or more, and near figures historical data before the EMA period.
are obtained for WMA with periods of 3 to 19. Those results Thus, this final residual weight is the residual from the
reveal a new disadvantage of the WMA: its weighting technique residual from the residual and so on, which is equal to (1-k) from
seems to be overweighted and biased significantly to the recent (1-k) from (1-k) and so on for (n) loops. Therefore, it is concluded
data. Moreover, there are no other weighting options between that we can calculate the weight of the historical data in the
the unweighted SMA technique and the overweighted WMA EMA by the equation:
technique. W(H) = (1-k)^n = [ 1- ( 2/(n+1) ) ]^n
Where W(H) is the weight of the historical data before the
Digging Deeper Into EMA Misses (n) period of the EMA.
In this section, we will explain in detail the EMA using the By this equation, we can plot a curve displaying the weights
methodology of applying logical questions and showing the of the historical data before the period in an EMA against
surprising answers that will all together expose the fragility of all periods of EMA from 2 to 200; the resulting data showed
the EMA technique. that the historical data always has 13.5% weight of the EMA.
Moreover, the curve turned out to look like an exponential
Question 1: What does the (K) factor in the EMA equation curve that may be the origin of the name of the average. This is
stands for? illustrated in Table 2 and Figure 2.
For the WMA of the period (n):
The weight of the current datum (highest weight of the most
recent datum) is = n / Σ weights
The weight of the datum previous to most recent datum is =
(n-1) / Σ weights
The weight of the oldest datum (smallest weight) is = 1 / Σ
weights.

PAGE 56 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Table 2. The weight of the historical data before the Table 3. The relation between the weight of current
period in an EMA datum and the weight of historical data in different
periods of EMA
Period (n) of EMA W(H) = (1-k)^n = (1-(2/(n+1)))^n
Period (n) Weight of current Weight of historical Ratio
2 11%
of EMA datum W(0) data W(H) W(H):W(0)
3 12.5%
13 14.28% 13.48% <
4 13%
14 13.33% 13.49% >
14 13.4%
20 9.52% 13.51% 1.42 times
17 13.5%
50 3.92% 13.53% 3.45 times
200 13.53%
100 1.98% 13.53% 6.83 times
200 0.99% 13.53% 13.66 times
Figure 2. The weight of the historical data before the
period in different EMA Question 3: From where and how are the 13.5% weight of the
historical data obtained from the weights of the data in the
EMA period?
For the WMA, the weight of the current datum is the highest,
and then the weights decrease regularly (with fixed value =
weight of the oldest datum) until the lowest weight of the oldest
datum to cumulatively form 100% of the WMA weights. While
for the EMA, the weight of the current datum is the highest
Knowing the weight of the historical data as a percentage (= weight of current datum in WMA), and then the weights
is a good issue, but the most important is knowing its relative decrease irregularly (with increasing rate then decreasing rate)
weight to other data weights in the average. To start, we have until the lowest weight of the oldest datum to cumulatively form
to standardize an allocation system for the data in the average, 86.46% of the EMA weights.
so we will consider the location (L) of the most recent current For the EMA, the weights of the data in the period show
datum as “0” and the previous datum as “1” and so on till the chronological decreasing rate, as illustrated in Figure 3.
oldest datum in the period having a location of “n-1”. Therefore,
we can calculate the weight (W) of any location (L) in a WMA Figure 3. The irregular decrease in weights through
locations in the EMA
with any period (n) by the following equation:
W(L) in WMA(n) = [ 2* (n-L) ] / [ n*(n+1) ]
And for the EMA: W(L) = K * (1-k)^L and since 1-k = (n+1 / n+1) – (2
/ n+1) = n-1 / n+1 so W(L) in EMA(n) = (2/n+1) * [ (n-1/n+1)^L ] = (2/
n+1) * [ (n-1)^L / (n+1)^L ] so finally:
W(L) in EMA(n) = [ 2 * ( (n-1)^L ) ] / [ (n+1)^(L+1) ]
By applying these equations to different locations and the
historical data in different EMA of different periods, we will find
surprising drawbacks in the weighting system of the EMA as
following:
1. The weight of the historical data before the period may be 1. During the recent 36% of the data (from the most recent
more than the weight of the oldest datum in the period; this current datum backwards to third the data), a rapid decrease
occurs very early in the period of (4): in weights occurs to form cumulatively 52.29% of average
EMA(3): W(0)=50%, W(1)=25%, W(2)=12.5%, W(H)=12.5% so weights. Also the weights are less than their similar locations
W(H)=W(oldest). in the WMA where the difference reachs its maximum at the
EMA(4): W(0)=40%, W(1)=24%, W(2)=14.4%, W(3)=8.64%, end of this partition.
W(H)=12.96% so W(H)>W(oldest). 2. From 37% to 79% of the data (middle 43% of data), a slow
2. Moreover, the weight of the historical data before the period decrease in weights occurs. At the end of this portion, the
may be more than the weight of the most recent (most cumulative weight reaches 79.81% of the average weights.
important) current datum in the period; this occurs early in Also the weights are less than their similar locations in the
the period of (14). The relation between them in different EMA WMA, where at the end of this partition the cumulative
periods are illustrated in Table 3. weight is equal in EMA and WMA.
3. From 80% to the oldest datum (oldest 21% of data), a very slow
decrease in weights occurs. At the end of this portion, the
cumulative weight reaches 86.46% of the average (and 13.54%
is left to the historical data before the period). Also the
weights are more than their similar locations in the WMA.

IFTA.ORG PAGE 57
 IFTA JOURNAL 2021 EDITION

Therefore, it is obvious that the weight of the historical data The SMA can be officially named “moving arithmetic mean”,
(13.5%) are obtained irregularly from the weights of the data in and the WMA is best named as “weighted moving average”. As a
the period causing an irregular rate of decrease in the weights point of view, it is more apt to describe the EMA as “exponential
of the data in the period. In fact, this is because the decrease smoother” rather than “exponential moving average”.
in the weights is by a regular percentage rate not a value rate,
where this percentage rate of decrease is equal to (K). Therefore, Question 6: What does the EMA actually perform on every data
each location weight is less than the previous location weight feed?
by K% of that previous location weight. So, EMA is a percentage The EMA does not deserve the name average, as it acts only
smoothing average. on the current new feed by giving it (K) weight, and it does not
consider averaging it with the data in the period preceding
Question 4: How far does the EMA go in history? it (depending on being included in the previous EMA). So the
Theoretically and mathematically, the SMA for the oldest residual weight is easily directed toward the previous EMA, and
period in the early beginning of the data has a calculated weight so on through the entire data.
in the recent EMA occurring after many decades of data. But For the SMA and the WMA, with every new feed, the oldest
practically speaking, the weight of the EMA is divided into datum in the previous average is neglected, and the new feed is
86.46% to the recent period and 13.54% to all the previous data, averaged again with the other data on a statistical concept of
where this 13.54%, in turn, is divided into 86.46% of it to the arithmetic mean or weighted mean, so the non-used portion of
previous period, and 13.54% of it is left to the data before the the current datum is dismissed due to the averaging, which has
recent two periods, and so on. The cumulative weight of periods a tremendous psychological consideration.
of data in the EMA is illustrated in Table 4. However, EMA is not an average; therefore, we will look to it
from another view than the weights in an average. So with every
Table 4. The cumulative weight of previous periods in new feed, the vertical amplitude between the previous EMA and
EMA
the current new feed is calculated, the new EMA moves in the
Previous periods of data Cumulative weight in current EMA direction of the new datum by (K) percentage of this distance,
1 86.46%
and the residual of the distance is thrown with no statistical
bases and without any psychological rational. The user may not
1.5 95.02% understand that the EMA does nothing except slow the entire
2 98.17% chart by following the data line and covering only a certain
2.3 99.01%
percentage of the lag in each step.

3 99.75% Question 7: When does the EMA deflect?

3.5 99.91% Since the calculation of the current EMA takes a certain
4.5 99.99%
percentage from the current datum and the remaining
percentage from the previous EMA, by logic, the current EMA
This result uncovers the fact that the EMA considers (nearly must be located between the current datum and the previous
only) the recent two and third periods of data and does not EMA and can not break beyond the current datum or the
handle the aggregate historical data. previous EMA. Accordingly, when the current datum is higher
than the previous EMA, the current EMA will rise higher than
Question 5: What does the EMA actually perform in the period? the previous EMA, disregarding if the current datum is rising or
The EMA does not perform any specific action on the data in in the same level or declining relative to its previous datum.
the period other than that on the data before the period. So in Now we can conclude that the rising EMA will never reverse
fact, the EMA does not respect the period at all. It seems that the downwards except when crossed by its data to the downside
EMA misleads the user by entering a value in a data field called (i.e., the current data is lower than the previous EMA). Then,
“EMA Period”; while this value will not be used as its name, the the current EMA will turn to decrease the downward distance
EMA uses the period value only to create a smoothing factor between the EMA line and the data line (the EMA line reverses
(K) simulating the first weight in the WMA with a period like only after crossed over by its data line). Unlike the smart SMA
the entered value. Afterwards, the EMA uses this generated or WMA, when the current datum is higher than the previous
“constant” to smooth all the data, disregarding the data in the average, but the current datum is much lower than its previous
period from the data before the period and giving no advantage datum, this extreme value (especially in WMA) will significantly
to the data in the period. affect the current average, so that the current average may
Therefore, the EMA really does not “average” the data in a be lower than the previous average. Therefore, the moving
specific period like SMA and WMA but smoothes all the data. average reversal can occur before the occurrence of actual (late)
Also, each SMA or WMA at a new point in time considers a new crossover of the current datum below the previous EMA. Finally,
datum and neglects an old datum, so actually it “moves” through we can describe the EMA as a slow follower, which resembles an
the data. However, each EMA at any point in time smoothes idiot slowly dog walking towards his master who rides a horse.
the all the data before it, so it is misnamed as “moving” and as All these answers now uncover the major mathematical and
“average” too. technical disadvantages of the EMA:

PAGE 58 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

1. It considers almost only an additional 1.3 period of the Advantages of CMA

historical data. 1. When using the simple averaging method of CMA and
2. This historical data has a weight more than the weight of the selecting a coefficient zero, this simulates the SMA, as all
current datum in periods greater than 13. data have equal weights, but with the edge of including the
3. The decrease in weights of the data through the period is historical data.
irregular. 2. When using the simple averaging method of CMA and
4. It does not consider the period by any specific action. selecting a coefficient one, this simulates the WMA, but with
5. It is not an average to the period but a smoother to the entire the edge of including the historical data. Taking advantage
data. over the simple CMA with coefficient zero as being weighted
6. It is a slow follower (reverse only after crossed over by its and affected by extreme values only upon entering the
data line). average.
3. When using the simple averaging method of CMA and
Coefficient Moving Average (CMA) selecting a coefficient between zero and one, this breaks the
cuffs and opens the horizon for the user to customise his
Calculation and Concept of CMA own suitable regular tested weighting intervals between the
Two user-entered parameters are the data array to be unweighted SMA and the overweighted WMA.
averaged (D) and the period of averaging (n). Each average value 4. CMA is a real average that not only greatly considers the
will act on (n+1) number of data, arranged as follows: the CMA of averaged period but also includes historical data in the form
the previous period, then the oldest datum in the period. Then, of the averaged previous period value, then calculates the
proceed chronologically forward until the most recent current statistically logical arithmetic or weighted mean. The previous
datum. Other two user-entered parameters are the value of average also contains the before previous average and so
averaging coefficient (C), ranging from “0” to “1”, and the type of on, so CMA is smart to track its own way backwards in the
averaging (T), which is either simple “0” or compound “1”. historical data by including only its average before the period.
• The CMA of the previous period is given a weight of (1). 5. Regarding the depth of the significant considered history,
• When simple averaging is selected, summation is used: weight at least one previous period from the historical data is
of each datum = weight of previous datum + coefficient. guaranteed, as it is already logically embedded in the
• When compound averaging is selected, multiplication is averaging mathematical equation, while theoretically all the
used: weight of each datum = weight of previous datum * ( 1 + historical data are included.
coefficient ). 6. In addition to including the historical data, CMA ensures:
• The current CMA is calculated like the WMA: Current CMA = a) The decrease in weights of data as we go backwards is not
summation of weighted data / summation of weights. only regular but also user-customised.
b) The weight of the historical data is definitely less than the
The weighting system is clarified by an example in Table 5. weight of the oldest datum in any averaged period.
c) A continuous regular dilation of weights is performed
Table 5. Example for CMA weighting system through the data in the averaged period, and then through
Data (D) = C “closing price”, Period (n) = 5, Coefficient (C) = 0.5 the historical data.
7. The compound averaging method of the CMA allowed a regular
1 5
CMA of (oldest (current percentage weighting system (not as the regular value weighting
previous datum in datum in Total system of the simple CMA) that also averaged a considered
period period) 2 3 4 period) weights
period and added the historical data with the least weight.
Simple 1 1.5 2 2.5 3 3.5 13.5
CMA
(C,5,0.5,0)
CMA Amibroker Formula Language
(AFL) Code
Compound 1 1.5 2.25 3.375 5.063 7.594 20.782
CMA
(C,5,0.5,1) Indicator Inputs
AveragedData = ParamField(“Data Field”,3);
At the beginning of the data: Period = Param(“Period”, 14, 2, 300, 1, 14);
In the oldest period, from the oldest datum until the datum Coefficient = Param(“Weight Coefficient”, 0.5, 0, 2, 0.01, 0.1);
before the most recent, there is no calculated CMA. WeightType = ParamToggle(“Weight Type”, “Simple|Compound”, 0);
In the period after the oldest period, starting from the most
recent datum in the first period until the datum before the most Calculate Weights and Total Weights
recent datum in the second period, the CMA is calculated for Weight[0] = 0; TotalWeights = 0; TotalWeightsWithoutHistory = 0;
(n) data (not n +1)—i.e., without previous CMA, with the oldest for ( Multiplier = 0; Multiplier <= Period; Multiplier++ )
datum in the period taking the weight of (1), and the coefficient { if ( (WeightType) == 0 ) Weight[Multiplier] = 1 + ( Coefficient *
augments the weights through the period. Multiplier );
Starting from the most recent datum in the second period if ( (WeightType) == 1 ) Weight[Multiplier] = ( 1 + Coefficient )
until the end of the data, CMA is calculated normally. ^Multiplier;
TotalWeights = TotalWeights + Weight[Multiplier];

IFTA.ORG PAGE 59
 IFTA JOURNAL 2021 EDITION

if ( Multiplier == Period ) TotalWeightsWithoutHistory = AveragedData[BarPosition+AveragedBars] * Weight[Counter] );

TotalWeights - Weight[Multiplier]; } Counter++; }
CoefficientMovingAverage[BarPosition] =
Find the Start Position and Calculate First CMAs CummulativeWeightedData / TotalWeights; }
Without Historical Data
CoefficientMovingAverage[0] = 0; Plot CMA
for ( i = 0; i <= BarCount -1; i++ ) if ( AveragedData[i] > 0 OR Plot ( CoefficientMovingAverage, “CMA”, ParamColor( “CMA
AveragedData[i] < 0) { StartPosition = i; break; } Color”, colorCycle ), ParamStyle(“CMA Style”), 0, 0, 0 );
for ( BarPosition = StartPosition + Period -1; BarPosition <=
StartPosition + ( 2* ( Period -1 ) ); BarPosition++) Testing Results
{ CummulativeWeightedData = 0; Counter = 0; The testing of the CMA was performed on the stocks of
for ( AveragedBars = 1- Period; AveragedBars <= 0; the S&P 100 index for three years, from January 2016 to
AveragedBars++) December 2018 on the daily timeframe. The strategy used
{ CummulativeWeightedData = CummulativeWeightedData + ( was the simplest, which was the single crossover tactic; buy
AveragedData[BarPosition+AveragedBars] * Weight[Counter] ); when the closing price crosses the moving average upwards
Counter++; } (Bullish crossover), and sell when the closing price crosses the
CoefficientMovingAverage[BarPosition] = moving average downwards (Bearish crossover). Also, only long
CummulativeWeightedData / TotalWeightsWithoutHistory; } positions were taken for the sake of simplicity.
The portfolio was adjusted to allow the maximum of 100
Calculate the Following CMAs With Historical Data simultaneous open positions to allow open trades in all available
for ( BarPosition = StartPosition + ( 2* Period ) -1; BarPosition <= stocks at the same time. Besides, the position size for each trade
BarCount -1; BarPosition++) was set as 1% of the portfolio to allow equal exposure.
{ CummulativeWeightedData = CoefficientMovingAverage[BarP Different types of moving averages were tested using the
osition-Period]; Counter=1; closing price data, and the period of 19 days was selected (based
for ( AveragedBars = 1- Period; AveragedBars <= 0; on previous mapping for the best CMA period selection). Then,
AveragedBars++) the results were clarified in Table 6.
{ CummulativeWeightedData = CummulativeWeightedData + (

Table 6. Testing CMA against other averages on the S&P 100

CMA (0.45, CMA (0.1,

Item Simple) Compound) SMA WMA EMA
Whole Trades
Number of All Trades 94 92 92 129 89
Average Outcome Percentage per Trade 1.04% 0.92% -0.06% -0.56% 0.07%
Average Exposure Percentage 98% 98.7% 97.5% 97.6% 98.1%
Winning Trades
Number of Winning Trades 30 29 26 32 24
Average Profit Percentage per Winning Trade 6.19% 6.4% 4.39% 3.01% 5%
Average Bars Held per Winning Trade 19.9 21.2 20.3 15.8 23
Maximum Consecutive Winning Trades 3 4 2 4 3
Losing Trades
Number of Losing Trades 64 63 66 97 65
Average Loss Percentage per Losing Trade -1.37% -1.61% -1.81% -1.74% -1.75%
Average Bars Held per Losing Trade 3.7 3.5 4.5 3.7 4.3
Maximum Consecutive Losing Trades 10 7 7 19 11
System Return & Risk
Compounded Annual Return 32.57% 26.43% -3.61% -23.15% -0.74%
Maximum Trade Percentage Drawdown -7.9% -7.9% -8.1% -14.8% -11.4%
Maximum System Percentage Drawdown -16% -23.9% -28.2% -56.8% -26.5%
Ulcer Risk Index 5.73 8.59 8.57 31.35 9.4

PAGE 60 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

CMA (0.45, CMA (0.1,

Item Simple) Compound) SMA WMA EMA
System Risk/Reward Ratios
Profit Factor 1.89 1.62 0.91 0.55 0.98
Payoff Ratio 4.02 3.52 2.32 1.66 2.66
Probability of Winning Trade 31.9% 31.5% 28.3% 24.8% 27%
Compounded Annual Return / Maximum System
2.03 1.11 -0.13 -0.41 -0.03
Percentage Drawdown
Ulcer Performance Index 4.81 2.49 -1.01 -0.9 -0.61
K-Ratio 0.12 0.09 0.02 -0.04 0.04

Figure 4. UPI of CMA with different periods and

Discussion coefficients on the S&P 100

CMA Combined With Time Analysis

The technical analyst identifies an active cycle in the
prices when a three or more successive waves (move and its
correction) take nearly the same duration. Therefore, it is
mandatory to adjust the period of the moving average equal to
this cycle duration. However, many times, the simple moving
average (before backward shifting) lies far below some price
bottoms, and some other price bottoms false break this moving
average. Moreover, it is common that the channels around the
centered moving average can barely contain the price peaks and
bottoms accurately without false breaks or being far from price
turning points.
An ultimate solution is to use the CMA that is not only Figure 4 represents a 3D chart where the X-axis shows the
committed to the period and averages only the prices through different periods of the CMA ranging from 2 to 50; the Y-axis
the wave duration. But also, the CMA coefficient has a wide shows the different coefficients of the CMA ranging from 0 to 1
range to be fine-tuned, allowing the CMA line to convincingly (with intervals of 0.05); the Z-axis shows the Ulcer Performance
touch the price bottoms (before backward shifting), and leading Index (with a water level at 0.5 to facilitate visual analysis). This
the channels to accurately hug the price turning points (after chart proves the following:
centering the CMA). Therefore, the CMA opens the limits for the 1. Some CMAs, with coefficients between zero and one, have UPI
time analysts to enhance their use of channels in time cycles. not only higher than the UPI of the CMA with the same period
Coefficient adjustment enhance Risk-adjusted Return: and coefficient zero (that resemble the SMA with considering
In the past, we were blocked to use only the unweighted SMA the history), but also higher than the UPI of the CMA with
or the overweighted WMA, but the CMA opens the wide range of the same period and coefficient one (that resemble the WMA
weighting system between both of them. To know whether that considering the history).
adjustable coefficient is beneficial, we studied the CMA over the 2. This situation is not a single spike in the chart but a chain
constituents of the S&P 100 by the single crossover tactic for of mountains as occurring in the periods from 17 to 27 with
three years starting from 2016 until 2018. nearly all the coefficients between zero and one. This proves
Ulcer Risk Index = Square root of [ Sum of Squared System that it is not an exceptional case found by a wicked data
Percentage Drawdowns / Number of Bars ] mining, but it is the fact that we were in a bad need to open
Unlike the standard deviation, the ulcer risk index measures the field of weighting systems to enhance the moving average
only the downward volatility, and from the previous peak strategies.
rather than from the mean. Therefore, it represents how much
the system can maintain achieved returns, by less in number
and shallow (depth of drawdown) and rapidly recoverable
drawdowns (duration of drawdown).
Ulcer Performance Index (UPI) or Martin ratio = [ Total Return
- Risk Free Return ] / Ulcer Risk Index
UPI measures the risk-adjusted return of the equity curve of
the trading strategy.

IFTA.ORG PAGE 61
 IFTA JOURNAL 2021 EDITION

Conclusion • CMA with the adjustable coefficient between zero and one
• The EMA has a lot of statistical and logical pitfalls that make allows the user to apply his own suitable tested weights;
its use questionable and should be revised. also the compound coefficient introduces a new edge in the
• The averages used in the equations of many technical weighting system.
indicators are logically to be arithmetic average, but SMA • CMA with the adjustable coefficient allows better time
is usually avoided because it can give zero value that ruins analysis using channels around the centered moving average.
the equation if the average value is in the denominator
of a division. However, it is advised that the CMA with a References
coefficient zero replace all these averages, as it resembles the Elder, Alexander, The New Trading for a Living. John Wiley & Sons, 2014.
SMA but with the advantage of containing the historical data, Kaufman, Perry, Trading Systems and Methods. 5th ed. John Wiley &
so it can rarely give a zero value. Sons, 2013.
• The averages used as a technical tool in a trading strategy are
Kirkpatrick, Charles D., and Julie A. Dahlquist. Technical Analysis: The
advised to be the logical WMA, as it has a statistical concept Complete Resource for Financial Market Technicians. 3rd ed., FT Press,
in averaging, with the advantage of weighting the recent 2016.
data. However, it is advised that the CMA with a coefficient
one replace all these averages, as it resembles the WMA but Software and Data
with the advantage of containing the historical data, and it Testing was performed by AmiBroker software, version 6.2.1.
can give better results. Data were provided by Thomson Reuters data feed.

PAGE 62 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Mohamed Ashraf Mahfouz, CETA, CFTe, MFTA

The Ripples Effect: A Clearer View for c/o Commercial International Brokerage Company (CIBC)
West Tower, Galleria 40,
Market Action and Price Patterns 26 of July axis extension Sheikh Zayed,
6 of October City, Egypt
By Mohamed Ashraf Mahfouz, CETA, CFTe, MFTA +2 01001842420
[email protected]

Chart 1. Ezz Steel Company (ESRS.CA) weekly candlestick

Introduction chart with volume
How do markets move? It is question that is asked by
everyone, but unfortunately it has no clear or confident answer.
For example, a fundamentalist would say that markets move
randomly around its intrinsic value, without denying the
trending manner the markets move in. A technical analyst or
chartist would say that markets move in trends, surely without
denying the existence of randomness, or doubtful areas, and
the continuous false breaks the markets witness during these
trends. Can it be inferred, then, that markets move in clear
trends, but in a random way? This article tries to answer this
question by discussing the causes for market movements. It
tries to analyze how to identify the general market trend and
expect the random movements that constitute it using a simple As can be seen in Chart 1, the stock was in a downtrend
logical tool. The article draws heavily on the work of Brian J. from Q4 2014 until Q3 2016, forming consecutive lower highs
Millard (1999) on channels. and lower lows and showing that bears were in control. In Q4
2016, the stock achieved a clear higher low formation as it did
Price Movement: Questioning the not make a new low and started to rise steeply, breaking the
Demand and Supply Equation previous peak to the upside and was accompanied with high
When it comes to market action it is always seen as a result volume. At this point, this kind of breakout indicates that bulls
of the demand and supply equation or bulls versus bears. If are in control and the next direction is going to be to the upside.
bulls are stronger than bears the trend is believed to be up and That is exactly what happened afterwards, as shown in Chart 2.
consists of continuous higher lows and higher highs. On the
other hand, if bears are stronger than bulls the trend is then Chart 2. Ezz Steel Company (ESRS.CA) weekly
candlestick chart with volume (update)
considered down and consists of continuous lower highs and
lower lows. Sometimes if volume is available we can use it to
confirm those directions or breakouts.
Based on the previous logic, price patterns are defined as
graphical formations that appear on the price chart, have
predictive value, and are created by demand and supply forces.
Accordingly, if we have, for example, a horizontal resistance
level and a higher low pattern, we identify it as a bullish
ascending triangle, and the next direction bias will be to the
upside. On the other hand, if we identify a head and shoulders
pattern thatconsists of same lows between three peaks, but the
third peak is a lower high (after a higher high), the next direction
biasedness will be to the downside etc.
The following charts of Ezz Steel Company (ESRS.CA) in But does it always work like this?
the Egyptian Stock Market are showing the above-mentioned
situation as follows: The previous example proves that the market direction is
indicated by the supply and demand forces represented by peaks
and bottoms. Unfortunately, this is not totally true, as there are
lots of other examples that show the same situation, but the end
result or direction is different. See the following examples.

IFTA.ORG PAGE 63
 IFTA JOURNAL 2021 EDITION

Example 1: Example 2:
Chart 3. Qalaa Holding Company (CCAP.CA) weekly Chart 5. The Egyptian Stock Market index (.EGX30) daily
candlestick chart with volume candlestick chart

As can be seen in the Chart 3, the stock was in a downtrend Chart 5 displays the EGX30 daily chart; it shows three
from Q4 2009 until Q2 2012, forming consecutive lower highs different directions—two downtrends and one uptrend.
and lower lows, showing that bears were in control. In the Q3 There are three circles on the chart that highlight the break of
2012, the stock achieved a clear same low formation, as it did not resistance levels to the upside after a higher low formation. This
make a new low and started to rise, breaking the previous peak situation should always indicate a trend reversal, or at least a
to the upside and was accompanied with high volume forming bullish scenario. Despite those upside breakouts, only one of
a double bottom formation. At this point, this kind of breakout them resulted to the upside, and the other two resulted to the
indicates that bulls are in control and the next direction is going downside. Such situations are referred to as failed breakouts,
to be to the upside. But this time, that did not happen, and the whipsaws, or bull traps because, according to the demand
stock, after a minor rise, started to decline, as shown in the and supply analysis, simply nothing can justify that an upside
Chart 4. breakout will result in a downward direction.
Chart 4. Qalaa Holding Company (CCAP.CA) weekly Example 3:
candlestick chart with volume (update)
Chart 6. The Dow Jones Industrial Average (.DJI) weekly
candlestick chart

Thus, a new high breakout in this case is considered

misleading, and the stock went down. From demand and supply
Chart 7. The Dow Jones Industrial Average (.DJI) weekly
perspective, this case does not make any sense; to break a candlestick chart (update)
resistance with high volumes and the stock still goes down
is very weird and questions the whole theory of demand and
supply.
The following are examples of two indices of the Egyptian
Stock Market and the U.S. market.

Charts 6 and 7 demonstrate the Dow Jones Industrial Average

(.DJI) in year 2010. The index has achieved a clear head and
shoulders pattern on the weekly chart during the first half of
2010, with a clear breakout to the downside.

PAGE 64 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

As can be seen in Chart 7, despite this clear downside take profits, but what if the time passed and the stock did not
breakout, the end result was to the upside, and the index went rise? The trader will still sell it even if he realized a loss because
to new highs as if nothing happened in the first half of 2010; the he is a day trader (time). The same characteristics of the day
only commentary or justification will be that the pattern failed. trader are applied on the short-term investors that invest during
There are many examples that show bearish patterns or three weeks or one month duration and are also applied on the
signals that resulted in bull trends or positive direction and medium- or long-term investors that invest in the market every
bullish patterns or signals that resulted in bear trends or year or three to five years or more.
negative direction. Thus, the market action should be looked The following examples will show six different situations of
at from a different perspective besides demand and supply how investors with different time spans act together to create
(bottoms and peaks). the market trend using the S&P 500 index daily price chart.

Controversy Over Market Example 1:

Movements When nearly all types of investors are acting positively inside
The main reason why the market it is looked at in terms of the market.
demand and supply is that people want to justify the force
behind the market upside and downside movements. But if the Chart 8. The S&P 500 index (.SPX) daily candlestick chart
main forces are demand and supply, then the economic theory
would imply that markets are moving around an equilibrium
point (demand=supply). So if the demand is bigger than the
supply, the price will increase, then the quantity of supply will
start to increase, and hence, the market will start to decline
until it reaches the equilibrium point. On the other hand, if the
supply is bigger than the demand, the price will decrease, then
the quantity of demand will increase, and hence, the market will
start to rise until it reaches the equilibrium point. This idea is
very similar to the fundamental analysis explanation to the price
action, as it suggests that markets are moving randomly around
their intrinsic fair values. Thus, if this justification is completely
right, markets won’t have any trend; they will always move in As can be seen from Chart 8, the market rose almost every
sideways. But practically speaking, this situation cannot be true day continuously except for a few situations where day traders
at all; we have seen markets that have been in a trend for years, affected the close of some sessions, but they do not interrupt the
either to the upside or to the downside. That is why one of the direction.
main Dow Theory tenets says that “A trend is assumed to be in
effect until it gives definite signals that it has reversed.” Example 2:
When medium- and long-term investors are acting positively
Main Idea inside the market but short-term investors are acting
The main reason why markets move in trends in a zigzag negatively.
way is that it is composed of more than one demand and supply
equation, or in other words the market consists of more than one Chart 9. The S&P 500 index (.SPX) daily candlestick chart
TIME Span direction that act together at the same point of time.
Dow Theory states that, “The Market has three trends”. Dow
considered a trend to have three parts; primary, secondary and
minor which he compared to the tide, waves and ripples of the
sea. This explanation is very true about how the market really
acts. It should be emphasized however that these three parts
should not be considered as three different stages. We believe
that the three trends are acting together at the same time all
the time. Thus, instead of dividing the trend into bulls and bears
or long-term versus medium-term trends one should divide it
into long, medium, and short-term investors acting inside the By looking at the chart above, it is very obvious how the
market at the same time. This was explained by the time cycle short-term investors’ negative attitude affected the market
analysis in the sense that the trend consists of all active cycles direction in some situations. Eventually, the end result is that
or what is referred to as the summation principle. the market will continue to go up as the medium- and long-term
We believe that the main drive for investment is time before investors are still positive. That is why the market will witness
the amplitude. To clear this point up, let us assume that we have lots of negative sell signals, but it will resume to the upside, and
a day trader who entered a trade by buying a stock at a specific if there is a bearish pattern appeared it will fail.
level. If the stock rose during the session, this trader will sell to

IFTA.ORG PAGE 65
 IFTA JOURNAL 2021 EDITION

Example 3: Example 5:
When long-term investors are acting positively inside the When medium- and long-term investors are acting negatively
market but medium- and short-term investors are acting inside the market but short-term investors are acting positively.
negatively.
Chart 12: The S&P 500 index (.SPX) daily candlestick
Chart 10: The S&P 500 index (.SPX) daily candlestick chart
chart

It is very obvious from looking at Chart 12 how the short-term

It can be shown from the above chart how the market trend investors’ positive attitude affected the market direction in
can be interrupted for a whole year (Year 2015–inside the some situations. Eventually, the end result is that the market
circle). Despite the medium- and short-term selling pressure, continued to go down as the medium- and long-term investors
the overall trend direction did not shift from up to down, but are still bearish. That is why the market will witness lots of
it nearly moved sideways, and any downside breakouts failed positive buy signals, but it will resume to the downside, and if a
to put the market lower because the long-term investors were bullish pattern or upside breakout appears, it will fail.
buying the market.
This point is extremely important, and it must be clear that Example 6:
the next potential direction of any breakout of a support/ When long-term investors are acting negatively inside the
resistance or price pattern boundary is not based on its shape market but medium- and short-term investors are acting
or name but on the direction of the longer term trend. Thus, it positively.
is not based on demand and supply situation; it is based on the
direction of the different time span trends, and thus, the bigger Chart 13. The S&P 500 index (.SPX) daily candlestick
chart
time span trend will identify the next direction.

Example 4:
When nearly all types of investors are acting negatively inside
the market.
Chart 11. The S&P 500 index (.SPX) daily candlestick
chart

In 2017, the market was moving in a high volatility sideways

that contained more than a higher low formation and upside
breakouts, but they all failed to continue the upside movements.
Despite the medium- and short-term buying power, the overall
trend direction did not continue to the upside but rather shifted
from up to down after a medium-term sideways struggle
that ended to the downside with the power of the long-term
investors that were selling the market. Thus, the bigger time
span identified the next direction.
As can be seen from Chart 11, the market lost nearly 20
percent of its value in a continuous direction without any Price Ripples
interruption during December 2018. By examining the previous six examples, we will figure out
that the more all types of investors are in harmony taking the
same direction, the less fluctuation the market will witness.
On the other hand, less harmony between different types
of investors leads to more fluctuation in the market. This
fluctuation is created when two or more types of investors are

PAGE 66 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

moving against each other, which can be called price ripples or Hypothetical Illustration of the Effect of a High
price patterns. Time Span Trend on the Smaller One
Price patterns can now have a new definition: price ripples First, let us assume that we have only short-term investors in
that are created because of the collision of two or more different the market who buy the market on short-term equal intervals
trends (different in time and direction), the following direction of time. The expected end result is that the market will move in
of any price pattern is based on the direction of the longer term sideways movement with a relatively small magnitude. It will
trend, not on the structure or shape of the pattern. take the following shape:
Price patterns are similar to the water ripples that are created
when two or more water flows are moving against each other, as Figure 2. Market movement when there are only short-
term investors
shown in Figure 1.
Figure 1. Water ripples created as a result of the collision
of different direction water waves
Now let us assume that we have only medium-term investors
in the market who buy the market on medium-term equal
intervals of time, which is surely longer than that of the short-
term investors. The expected end result is that the market will
move in sideways movement but this time with relatively bigger
magnitude than that of the short-term investors. It will take the
following shape:
Figure 3. Market movement when there are only
medium-term investors

Chart 14 shows how the Ripples Effect appears on a real price

chart when two or more types of investors (in direction and
time) collide.
Chart 14. Emaar Misr for development of daily Finally, assuming only the existence of long-term investors:
candlestick chart
Figure 4. Market movement when there are only long-
term investors

If we hypothetically look at the long-term investors at the

same period of time that we examined the medium and short-
term investors, it will nearly create only one rise and one fall.
The previous three figures showed the hypothetical shape
of the market when we separate each type of investor; we
can clearly see that the end result of each is that the market
will move sideways, and the only difference is that the time
How are these ripples created? This question will be answered between bottoms will differ and of course the magnitude of
in the following section that explains how the trend is composed every rise will differ. Longer term investors will have greater
and how different time span trends affect each other. magnitude and more time elapsed between each successive
bottom. This is what is referred to as “proportionality
Identifying the Trend Composition principle” in Time Cycles Theory.
To try to understand how the trend moves when all types of After imagining how the market will move if it only includes
trends work together, it will be easier to illustrate first how each one type of investors, Figure 5 will show how the trend will
trend hypothetically works alone and then how the trend of a differ when we add more than one type of investor to the other
higher time span will affect the trend of a smaller time span. or what is referred to as the “summation” principle in Time
Cycles Theory.

IFTA.ORG PAGE 67
 IFTA JOURNAL 2021 EDITION

Figure 5. Adding the long-term investor to the medium-

term investor identify the different types of investors in real prices in order to
expect the next market movement. Figure 6 includes all types of
investors together and how the trend will be created.
Figure 6. Putting all types of investors together

As can be seen in Figure 5, we can divide it into three stages:

The first stage: The medium-term investor was affected
by the left side or the rising side of the long-term investor, so
whenever the medium-term investor is declining or pushing the Identifying Different Types of Investors in Price
market lower, it will not reach the previous low like when it was Charts
alone and will make a higher low, and when it rises, it will make The big challenge is to isolate different types of investors
a new high, not the same high as when it was alone. as we did previously, but this time on price charts. This can be
The second stage: This stage is very important when the long- done using the “moving average envelopes” or the “percentage
term investor starts to reverse from up to down (from left side envelopes” or the “envelopes” indicator and without the use of
to right side) and the medium-term investor is still involved and time cycle analysis.
going upward. This situation creates what we defined as price The envelopes are fixed bands that surround the moving
patterns or price ripples. In the above situation, the medium- average by a certain fixed percentage. It is originally made for
term investor nearly took the shape of the sideways as if it is the sake of filtration for the price crossover with the moving
alone. We can identify this sideways as a “double top” pattern. average during trending phases, and also made for trading
What makes us claim that it is a double top formation not a during sideways by widening the bands in order to trade
rectangle formation, or that in such case, the end result will between its boundaries.
be to the downside, is that the long-term investor will turn to In this article, we use the envelopes in a different way in order
the negative side, and the market won’t be able to break to the to isolate the different types of investors by enveloping the price
upside. Thus, the pattern is created by the intersection of both action of each investor separately. First, let us establish the
investors (long-term and medium-term), but what identifies the main rules when we apply envelopes.
next direction is the long-term investor. 1. Use the daily price chart only. All types of investors will
The third stage: The medium-term investor was affected by be identified from the same chart, including the long-term
the right side or the falling side of the long-term investor, so investors. The chart that will be used is always the shortest
whenever the medium-term investor is rising or pushing the term chart in our analysis, which is the daily chart.
market higher, it will not reach the previous high like when it For forex markets, we will use the hourly chart or the half-
was alone; it will make a lower high, and when it falls, it will hour chart as our main chart.
make a new low, not same low as when it was alone. 2. Use centered moving average envelopes. It is already known
By realizing the idea of those three stages, we can easily that envelopes are typically moving averages that are shifted
identify the direction of the investor of a longer time interval to the upside and to the downside. However they are averages
by looking at the peaks and bottoms of the investor of a shorter for the price, but they are plotted at the last traded price.
time interval. If, for example, the medium-term investor is They should, however, be plotted in the mid-point of the range
making a higher low, that means that the longer term investor under study. For example if we are using 10-day envelopes,
is still looking upward, preventing the medium-term investor every new point should be plotted on the fifth day not on the
to reach its previous low, and thus, we should be biased for a 10th day. In this way, the cycle for each type of investor will be
new high. And if the medium-term investor is getting weaker displayed in clearest way.
and is not able to make a new high, that means that the longer 3. Start drawing your envelopes from the small investor to
term investor is reversing. Lastly, if the medium-term investor the longer term investor. Thus, we will draw from the inside
is making a lower high and lower low, that definitely means that to the outside. Usually we draw three channels of envelopes
the longer term investor is declining. that represent the short-term, medium-term, and long-term
This idea is very important to understand before going investors. But you can use a longer term channel for secular
forward with our article because we will rely on it after we degree trends.

PAGE 68 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

The last general guideline is that you can use time analysis To solve this problem and try to avoid such false breaks, we
and price oscillators (Stochastic Slow Oscillator) to help you have to examine the different types of investors in order to be
choose the right span of the centered envelopes. able to predict the future direction of the stock based on the
The following section will present a case study by applying direction of each type of investor separately.
envelopes on real stock prices in order to identify different
types of investors in the same price chart. Step 1: Short-term investors
As we have mentioned before in the general rules, that will
Case Study: Emaar Misr for Development (EMFD.CA) in the draw our envelopes from the inside to the outside. First, we will
Egyptian stock market choose the envelopes that encompass price randomness. After
Charts 15 and 16 are daily price charts for Emaar Misr for choosing the right period and percentage of the envelopes to
Development, which is a stock in the Egyptian stock market. encompass the short-term investors’ fluctuation, remember
This chart was chosen as it nearly includes the three types of to center the envelopes in order to move identically with price
trends (Up – Sideways – Down). fluctuations.
Chart 15. Emaar Misr for development of daily Chart 17: Emaar Misr for development of daily
candlestick chart (Price Only) candlestick chart with centered five-day envelopes with
3.5% bands

Chart 16. Emaar Misr for development of daily

candlestick chart (with annotations)

In Chart 17, we can identify clearly repetitive bottoms and

peaks that are nearly equal or near in time, which shows
the short-term investors’ attitude or fluctuations. We can
also notice that despite the equally repetitive attitude of the
short-term investors, the trend was not sideways. Sometimes
it resulted in consecutive higher lows and higher highs, other
times to the downside, and other times sideways. That is
simply because this trend is not created just by the short-term
investors; there are longer time span investors that affected
the trend. This makes us move to the next step of our envelope
analysis.

Step 2: Medium-term investors

Looking at Chart 16, we can clearly see a lot of false breaks To identify the medium-term investors, we will choose the
in different situations and directions. We can see how many right envelopes that encompass peaks and bottoms of the
times the stock broke a support to the downside but rose smaller channel of the short-term investors. We will try as much
afterwards, and how it broke resistance to the upside but as we can to make the lower band touch most of the bottoms
declined afterwards. We can also see on the left side of Chart and the upper band touch most of the peaks of the short-term
16 a symmetrical triangle formation that was broken to the investors (smaller channel).
downside and should have acted as reversal, but it was a false
break or a failure, and the stock continued to move upward.
These situations happened in many cases and nearly in all
timeframes, especially in the small timeframes like daily, four
hours, hourly, and 30 minute charts.

IFTA.ORG PAGE 69
 IFTA JOURNAL 2021 EDITION

Chart 18. Emaar Misr for development of daily Chart 19. Emaar Misr for development of daily candlestick
candlestick chart with centered 5-day envelopes with chart with centered 5-day envelopes with 3.5% bands,
3.5% bands and centered 30-day envelopes with 7.5% centered 30-day envelopes with 7.5% bands, and centered
bands 100-day envelopes with 14% and 12.5% bands

Chart 18 shows both short-term investors (red channel) and After plotting the long-term channel, we can now
the medium-term investors (purple channel). The short-term understand the reason behind nearly every move within the
investors on the left side of the chart consist of consecutive trend. We can always justify the current move or direction of
higher lows and higher highs because there were longer term the trend understudy by understanding the direction of the
investors that were pushing the market higher. On the other longer term trend or investor. For example, on the left side of
hand, the short-term investors were making lower highs and the chart, since the biggest channel is moving upward, we can
lower lows on the right side of the chart because there were realize why every downside trigger failed and why, despite the
longer term investors that were pushing the market lower. In violation of the lower boundary of the symmetrical triangle—
cases where the short-term investors moved sideways, those which we referred to in chart 16—it did not continue to the
were the periods where the medium-term investors were downside and end to the upside.
creating a bottom or a peak within the trend. In the top of this chart, we can now understand that the
In Chart 18, the medium-term investor peaks and bottoms double-top formation was created because there were two
are now clear. We can also notice that despite the equally waves moving against each other; the long-term investors were
repetitive attitude of the medium-term investors, the trend reversing from up to down, and the medium-term investors
was not sideways. The left side of the chart above consists of were trying to move upward again, which caused the ripples.
consecutive higher lows and higher highs, and the right side Since the long-term investors were the party that is reversing
consists of consecutive lower highs and lower lows. That is to the downside, that meant for sure that this double-top
simply because this trend is not created only by the short-term formation will not fail and will result to the downside.
and medium-term investors but there are longer term investors Let us look again at the same chart after updating the last
that also contributed in this trend. This makes us move to the price.
last step of our envelope analysis.
Chart 20. Emaar Misr for development of daily candlestick
chart with centered 5-day envelopes with 3.5% bands,
Step 3: Long-term investors
centered 30-day envelopes with 7.5% bands, and centered
In order to identify the long-term investors, we will choose 100-day envelopes with 14% and 12.5% bands
the right envelopes that encompass peaks and bottoms of its
smaller channel of the medium-term investors. We will try
as much as we can to make the lower band touch most of the
bottoms and the upper band touch most of the peaks of the
medium-term investors (relatively smaller channel).

PAGE 70 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Chart 5. The Egyptian Stock Market index (.EGX30) daily

By examining the chart above and by looking at the recent candlestick chart
price action, ignoring the channels we drew, we will see that
the stock is moving above a strong support and is not violated
to the downside, as can be clearly seen in the drawn circle.
Also, if we try to focus on the short-term price action inside the
circle, we can see that the price did not break the support and
also, there are minor upside breakouts within this sideways
range. This strong support condition that also corresponds
to the major broken peak on the left side of the chart and the
mentioned minor upside breakouts could make us biased that
the end result of this situation will be to the upside, or at least to
continue sideways until we have more confirmation on the next
direction. But as we explained earlier, whatever the situation
and whatever the breakout direction, to anticipate the next We are going to apply the envelopes on the chart by drawing
two different time span channels to see what we get.
market movement, we have to look at the trend or the long-term
investor; thus, the current sideways movement (price ripples) Chart 22. The Egyptian Stock Market index (.EGX30)
is created because of the collision of more than one wave at daily candlestick chart with centered 17-day envelopes
the same time, and by looking at the long-term channel (green with 3.5% bands and centered 55-day envelopes with 6%
bands
channel), we can clearly see that it is still looking downward,
which means that the end result of the price ripples in the circle
will be to the downside, and that the strong support will be
violated to the downside. Please take a look at the update of
Chart 20 below in Chart 21.

Chart 21. Emaar Misr for development of daily

candlestick chart with centered 5-day envelopes with
3.5% bands, centered 30-day envelopes with 7.5% bands,
and centered 100-day envelopes with 14% and 12.5%
bands (updated prices)

In Chart 22, we can easily anticipate which upside breakout

will be successful when we examine the direction of the longer
term channel, when the longer term channel is rising; this is the
situation when any upside breakout will succeed, as it is moving
within the flow of the longer term wave.

Tools and Guidelines That Can Enhance Envelopes

Analysis
Like any tool, using envelopes in that manner has some
shortfalls. The main problem is that since envelopes are
centered moving averages, it will always lag the price action by
the amount of centralization. Thus, it is recommended to use
some guidelines while using it or to use another tool besides
As can be seen in Chart 21, the stock breached this strong using it to overcome this drawback.
support to the downside as expected because of the effect of the
long-term investors that were looking downward. Envelopes direction can be identified from the envelopes of
shorter time span
Case Study 2 A very important rule is that markets take time to reverse
their direction. In other words, envelopes will take time to
In this quick practice, Chart 5 of the Egyptian stock market
reverse from up to down or from down to up.
index EGX30 (shown below) is revisited to explain why, although
Thus, if we have envelopes that are looking downward and the
there were three upside breakouts, only one of them succeeded envelopes of shorter time span are moving upward—meaning
and the other two turned negative. two waves are moving against each other—we have to expect
that the smaller wave is the one that will be forced to look
downward again because of the wave of longer time span.

IFTA.ORG PAGE 71
 IFTA JOURNAL 2021 EDITION

Chart 24 Emaar Misr for development of daily

On the other hand, if the smaller wave started to go down but candlestick chart with centered 30-day envelopes with
it is not able to make a new low (higher low) then this definitely 7.5% bands and centered 100-day envelopes with 14%
means that the envelopes of longer time span will move at least and 12.5% bands (with a hypothetical expectation of the
sideways and could be reversing upward. This can be shown channels next direction)
from the example below.
Chart 23. Emaar Misr for development of daily
candlestick chart with centered 30-day envelopes with
7.5% bands and centered 100-day envelopes with 14%
and 12.5% bands

Chart 25 will show the real end result of the price action.
Chart 25. Emaar Misr for development of daily
candlestick chart with centered 30-day envelopes with
7.5% bands and centered 100-day envelopes with 14%
In Chart 23, along with two envelopes that represent two and 12.5% bands (real update)
different investors, looking at the red and green arrows, it
shows us the collision of the two waves. Whenever the shorter
term investors touch one of the boundaries of the longer term
investors, ripples are created, and the short-term movement
shifts its direction.
At Point 2, the smaller investors were looking down, creating
lower highs. But since the bigger investors were still looking
upward, the smaller channel was forced to look upward again by
the strength of the bigger channel.
At Point 1, we can see how the smaller channel created a
higher low, as shown by the green arrows which makes us
understand, even if the bigger channel is lagging behind, that it
is reversing its direction from down to up and hence preventing
the smaller channel from creating a new low. Thus, we should
expect that the smaller channel will eventually make a new Stochastic Slow Oscillator helps in identifying turning points
high. After understanding the relation between the different
At Point 3, we can see how the smaller channel created channels, we can use the Stochastic Slow Oscillator (9,5,3) to
a lower high, as shown by the red arrows, which makes us provide the trigger of the turning points through the crossovers
understand, even if the bigger channel is lagging behind, that it of %K to %D. We believe that the Stochastic Oscillator is the best
is reversing its direction from up to down and hence preventing tool to provide the turning point at the exact peak or bottom;
the smaller channel from creating a new high. Thus, we should thus, it will enhance your trading and solve the lagging problem
expect that the smaller channel will eventually make a new low. of the channels.
At Point 4, this is a good practice to understand how we can Chart 26 will show the turning points of the smallest channel.
expect the next market movement despite the delay of the If you want to anticipate the turning point of the bigger
channels because of the centralization. As we can see inside channels, use the bigger time span chart. So, if you are using the
the smaller channel, the prices were moving sideways biased to daily chart, use the weekly or monthly charts; if you are using
up, which gave us the feeling that the market will break to the hourly chart, use the 4-hour or daily or weekly chart.
upside. But by looking at the lagging bigger channel, it is still
looking downward; thus, we can imagine that if we extend its
boundaries, it will hit the smaller channel at some point to either
push it to a new low or at least a higher low. Thus, we should
expect that the stock will go down because of the direction of
the bigger channel. Chart 24 is the same as Chart 23 but includes
an imaginary extension for the channels and our expectation.

PAGE 72 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Chart 26. Gold 4-hour candlestick chart with centered 4 Chart 28. Gold daily candlestick chart along with
4-hour envelopes with 0.4% bands, centered 30 4-hour Stochastic Slow Oscillator (9,5,3)
envelopes with 0.8% bands, and centered 210 4-hour
envelopes with 2% bands along with Stochastic Slow
Oscillator (9,5,3)

We can see the clear negative divergence trigger provided by

the Stochastic Oscillator with the price action that anticipates
and confirms the weakness we saw earlier by the channels (in
the 4-hour chart).

Momentum indicator can show weakness or strength of the

channels
The envelopes are centered and shifted moving averages;
As can be seen from Chart 26, all bottoms and peaks of the thus, they may be lagging and they take some time to reverse
smallest channel (purple channel) correspond to the %K & %D from down to up or from up to down. Calculating a four-period
crossovers. momentum for the envelopes itself could provide us with an
To benefit from the Stochastic Slow Oscillator in the bigger early signal that the channel’s momentum is slowing down in its
channels, the following example will clear this point. direction and is about to reverse. The following example of the
U.S. Dollar versus the Japanese Yen will illustrate this tool.
Chart 27. Gold 4-hour candlestick chart with centered
30 4-hour envelopes with 0.8% bands and centered 210 Chart 29: U.S. Dollar/Japanese Yen daily candlestick
4-hour envelopes with 2% bands along with Stochastic chart with centered 34-day envelopes with 1.8% bands
Slow Oscillator (9,5,3) and centered 120-day envelopes with 3% bands along
with 4-day momentum of the 120-day envelopes

In the chart above, we have excluded the smallest channel,

and as you can clearly see, the Stochastic Slow Oscillator will not We can see from Chart 29 how nearly every turning peak or
help to anticipate the turning points of the bigger channels. bottom of the big channel (blue channel) was preceded by a
As can be shown in the circle in the chart above, the smaller divergence with the momentum indicator (in the lower panel). It
channel created a lower high formation, indicating that the was showing the slowdown in the momentum of the rise or fall,
bigger channel is turning from up to down, and that is why indicating that the rising channel will reverse to the downside,
we could have expected a new low in prices, which actually and when it is falling, it indicates that it was about to reverse to
happened. the upside.
We can also use the Stochastic Slow Oscillator but on the
longer term charts (daily chart in this example) to help us in Spotlight on IFTA Conference 2019
confirming the reversal of the bigger channel, as can be seen in Predictions
Chart 28. I had the honor to participate as a speaker at the IFTA
conference that was held in Cairo in October 2019. The topic of
my session was actually the topic of this article. To prove the
validity of the topic, I had provided predictions on the Egyptian
stock market and the U.S. equity market based on the envelopes
analysis as explained. It is my pleasure to show you our
prediction at that time and the update six months later.

IFTA.ORG PAGE 73
 IFTA JOURNAL 2021 EDITION

The Egyptian Stock Market Prediction The U.S. Stock Market Prediction
Chart 30. The Egyptian stock market index EGX30 daily Chart 32. The Dow Jones Industrial Average (DJI) daily
candlestick chart with centered 55-day envelopes with candlestick chart with centered 50-day envelopes
6% bands and centered 180-day envelopes with 13% with 7% bands and centered 200-day envelopes with
bands, and centered 600-day envelopes with 30% bands 13% bands along with 4-day momentum of the 200-day
envelopes

In October 2019 the DJI was at 27,200 points. As you can see
from the chart above, the Ripples Effect was created during two
complete years (2018–2019), which is—as mentioned before—
the result of the collision of two different investors. Also, by
In October 2019, EGX30 was almost at 15,000 points. We have looking at the 4 momentum of the bigger channel (long-term
explained before that we should make our analysis from inside investors), it is evident that the momentum of the rising long-
to outside, so by looking at the smaller channel, we will find out term upward envelopes is diminishing, which made us expect
that it was moving sideways and that the bigger channel (the that the end result of this Ripples Effect will be to the downside
purple channel) was looking downward; thus, our expectation as the long-term channel is about to reverse. Thus, we were
then was that the index should test the year low again at expecting that the Dow Jones Industrial Average is at least going
12,500–12,000 points and could extend to make a new low to retest 2019 lows at 22,000–21,500. Chart 33, six months later,
until it reaches the lower boundary of the long-term investors confirms that.
channel (orange channel). The following chart is an update six
months later. Chart 33. The Dow Jones Industrial Average (DJI) daily
candlestick chart (six-month update)
Chart 31. The Egyptian stock market index EGX30 daily
candlestick chart with centered 55-day envelopes with
6% bands and centered 180-day envelopes with 13%
bands and centered 600-day envelopes with 30% bands
(six-month update)

Six months later, the market actions confirmed the previous

expectations. More importantly, staying above 12,000 points
may not be considered as a bearish situation but high volatility
sideways. Being below this area would make most of the
investors look at the market as a bearish market, as it will
create major lower high and lower low pattern. Yet, since the
market reached the lower boundary of the long-term channel,
we expect that the market will witness a significant medium- to
long-term rally.

PAGE 74 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Conclusion References
Observing the Ripples Effect or the periods of price Millard, Brian J. Channels & Cycles: A Tribute to J.M. Hurst. Traders Press,
fluctuations brought us closer to understanding how the market Inc., 1999.
actually works. It made us look at the market action from the Murphy, John J. Technical Analysis of the Financial Markets: A
relation of different time span investors’ perspectives rather Comprehensive Guide to Trading Methods and Applications. New
than looking at it from the demand and supply forces side. York: New York Institute of Finance, 1999.
Identifying the different types of investors using envelopes Edwards, Robert D. and Magee, John and Bassetti, W.H.C., Technical
has the following merits: Analysis of Stock Trends. Boca Raton: Taylor & Francis Group, 2007.
1. Having a clearer view for the market action.
Kirkpatrick, Charles D., and Dahlquist, Julie R. The Complete Resource
2. Understanding how the participation of different time span for Financial Market Technicians. New Jersey: Pearson Education,
investors at the same time creates the market trend. Inc., 2007
3. Having a new understanding of price patterns and
Pring, Martin J., Technical Analysis Explained: The successful Investor’s
understanding that it is created when two or more different
Guide to Spotting Investment Trends and Turning Points, Fourth
types of investors (in direction and time) collide, and that is Edition, New York: McGraw-Hill, 2002.
why we called them price ripples.
4. Not to rely on the price pattern shape to anticipate the next
market movement but to rely on the direction of the longer Software and Data
term channel. Thomson Reuters Eikon
5. Analyzing the market using short-term channels can help us Metastock, Equis International, a Reuters company
in anticipating the direction of the longer term channels. MS Office, 2010 Microsoft office
Data provided by Refinitiv
Also, we have seen how we can use the Stochastic Slow
Oscillator to help us in identifying turning points of the small
channels and to use it on the bigger charts to anticipate the
reversal of the bigger channels. Finally, we explained how to
apply the momentum indicator on the envelopes itself to provide
us with more information concerning the weakness or strength
of the channels.
In the end, no one will ever have a definite view on the market
because the market is simply inefficient, and if we all know
where the market is going, it will correct itself and move in the
other direction. Thus, this article is considered as a trial to help
us look at the market in a clearer and more efficient way and
finally to understand the reason behind the false breaks and
randomness, or what we have called the Ripples Effect.

IFTA.ORG PAGE 75
 IFTA JOURNAL 2021 EDITION

Yukitoshi Higashino, MFTA

Forecasting Major World Indices on Ichimoku 742-18 Kayadamachi

Yachiyo-shi, Chiba-ken 276-0044
By Yukitoshi Higashino, MFTA Japan

+8190-3499-4252
[email protected]

Abstract • The middle-point of the price range is regarded as important.

Ichimoku is a technical analysis method developed by Goichi • The three basic components of the Ichimoku theory are the
Hosoda (1898–1982), a Japanese financial market journalist, “time” principle, “wave structure” principle, and the “price
through his many years of research in financial markets. Even level” principle.
30 years after Hosoda’s death, Ichimoku is still widely used by • One of the important traits of Ichimoku is its “time” study.
traders and investors as an effective tool for analyzing markets Most market players focus on price moves and tend to make
and trade with in Japan. Hosada’s grandson took over Ichimoku light of the time factor. In Ichimoku, while price moves are
and has a study meeting every month in Tokyo. important, the time factor is more important than the price
Many of you may particularly know about “kumo (cloud)”. factor; without a solid “time” study, one cannot have a true
However, although Ichimoku is starting to be popularized by an understanding of the markets.
increasing number of traders and investors around the world,
it does not seem to be utilized as widely and as effectively as Composition of the Ichimoku Chart
it is in Japan. The reason is because Ichimoku is an integrated Figure 1 is a daily chart of the S&P 500. Ichimoku consists
set of multi-faceted market analysis principles and techniques, of a candle chart and five lines. Each five lines often serves as
including price projection, time projection, wave analysis, and support or resistance.
more. The Conversion line is the middle-point of the high-low range
The wide variety of techniques and concepts included in in the last nine periods (including the current period), and the
the whole Ichimoku theory make it highly challenging to fully Base line is the middle-point of the high-low range in the last 26
master. And, obviously, there is a language issue. Japanese is periods (including the current period). They may look similar
a tricky language for many Westerners. Japanese vocabulary to moving averages, but they are different. They may be called
is very different from English, and Japanese grammatical “moving middle-points.” In Ichimoku, it is considered that
structure is also totally different than English, making middle-points represent better equilibrium points in the market
translation from Japanese into English quite difficult. It is than moving averages. In moving averages, regardless of the
especially challenging to find the right English translations of volatility or the price swing in a given period, the closing price
many Japanese words used in the original Ichimoku theory. is the only thing that is counted. That is, even if the price swings
Therefore I have made an attempt to use plain English words vary widely in a period, it is not reflected. Ichimoku dismisses
instead of being “loyal” to the Japanese words in the original this and uses the middle-point indicators, whichinstead reflect
theory. the whole price range in a given period.
The following content revises my presentation at the IFTA The Leading span 1 is the middle-point between the
Conference in Cairo and adds related theories to make it easier Conversion line and the Base line. The Leading span 2 is the
to understand for readers. I hope this will help IFTA colleagues middle-point of the high-low range in the last 52 periods
learn the basics of a unique technical analysis method developed (including the current period). Leading spans 1 and 2 are
in Japan. plotted 26 periods ahead (including the current period). The
cloud-like area formed by the Leading span 1 and 2 is called the
Introduction “Resistance band.”
• Ichimoku is focused on the underlying “powers” in the The lagging span is drawn by plotting the current closing
market. To know the next market direction, it suffices to price 26 periods backward (including the current period). It
know which side, buyers or sellers, is winning or losing. becomes a line which runs parallel to the current price line.
The market moves in the direction in which the equilibrium
between the buyers and sellers has been broken. The chart
developed by Hosoda allows one to instantly grasp the
equilibrium state of the market. This is why it was named
“Ichimoku Kinko Hyo,” which literally means “One Glance
Equilibrium Chart” in Japanese.

PAGE 76 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Figure
Figure1.1. Composition
Composition of theChart
of the Ichimoku Ichimoku chart Figure 3. Daily chart of 3M as asample of “San’yaku
Koten”
Figure 3. Daily chart of 3M as a sample of “San’yaku Koten”

Three Conditions to Make a Safe Bull

(Bear) Call The Three Basic Principles of
When the following three conditions are in place, one can Ichimoku
safely judge that the market is in a bullish state. We call it After confirming San’yaku Koten, you should consider the
“San’yaku Koten,” meaning the market is ready to start an three basic principles: “wave structure,” “price level,” and
uptrend. “time”. Figure 4 shows all the points. Notice that the bull market
• The Conversion line is above the Base line, which is trending consists of three waves, and an equilibrium is established
up or flat. between price and time. The same applies to the bear market. It
• The Lagging span is above the price of 26 periods ago. is highly probable that the market direction will reverse at the
• The price is above the Resistance band. equilibrium point of the horizontal and vertical axes.

On the other hand, when three opposite phenomena are in Figure

Figure 4.4. The
The three
three basicbasic principles
principles of Ichimoku
of Ichimoku

place, we call it “San’yaku Gyakuten,” meaning that the market

is ready to start a downtrend.
• The Conversion line is below the Base line, which is trending
down or flat.
• The Lagging span is below the price of 26 periods ago.
• The price is below the Resistance band.

Figure 2 is a daily chart of FB as a sample of “San’yaku Koten.”

As of (1), the Lagging span is above the price of 26 periods ago.
As of (2), the conversion line is above the Base line. As of (3), the
price is above the Resistance band. When these three conditions
are confirmed, you can judge “San’yaku Koten.”
Figure 2. Daily chart of FB as a sample of “San’yaku Wave Structure Principle
Koten”
Figure 2. Daily chart of FB as a sample of “San’yaku Koten”
First of all, the wave structure principle is very simple.
Hosoda classified the wave patterns that appear in financial
markets into a number of groups according to their wave
structure and gave them unique names. I wave is a single
rectilinear or straightish (normally sharp) thrust up or down
without notable corrective moves. V wave consists of two
successive I waves, a sharp thrust up followed by a sharp thrust
down, or a sharp thrust down followed by a sharp thrust up. N
wave is an up-down-up or down-up-down wave. This is the wave
pattern most commonly seen in the market.
When N wave extends to nine waves, it is judged that the
uptrend is close to the limit. And when the price falls below
the previous low, the uptrend terminates with the top-forming
Figure 3 is a daily chart of 3M. (1) to (3) are the same sign; pattern.
however as of (4), when the Lagging span is above the Resistance
band, the momentum is often stronger. And, without falling to
the baseline, the Conversion line provides continue to support
price as strong market.

IFTA.ORG PAGE 77
 IFTA JOURNAL 2021 EDITION

Figure
Figure 5.5. Wave
Wave structure
Structure Principleprinciple
Price Level Principle
In ichimoku, there are four basic projection methods as
shown below.
N projection -- Up: N = C + (B - A) Down: N = C - (A - B)
[Figure7]
* These two equations are effectively the same, but I am
showing both as I believe this makes it intuitively easier for
readers to understand for. The same applies to the following.

• Up: N projection adds the distance of the last upleg to the last
low.
• Down: N projection subtracts the distance of the last downleg
from the last high.
Figure 6 is a daily chart of the Nikkei Stock Average from Figure 7.
March 19, 2019, to January 28, 2020. Figure 7. NNProjection
projection
• The wave from A to B is an “I wave.” Up: N = C + (B - A) N projection Down: N = C - (A - B)

• The wave from B to C is also an “I wave.” N A

• The wave from A to C is a “V wave.”
• The wave from A to D is an “N wave.”
B C
• The wave from D to K is an “N wave” structured upward wave.
= =
The uptrend would terminate if the price falls below the low J.
Figure
Figure 6.6. Daily chart
of the of theStock
Nikkei stock average =
Daily chart Nikkei Average
= C B
I K
G
A N

V projection – Up: V = B + (B - C) Down: V = B - (C - B) [Figure 8]

A E J • Up: V projection adds the distance of the last downleg to the
H
C last high.
• Down: V projection subtracts the distance of the last upleg
from the last low.

F This is the target to sell half of the position.

B D Figure
A 8.
Figure 8. V Projection
V projection E

Up: V = B + (B- C) V projection Down: V = B - (C - B)

V A

B = C

= =
B
C =
A V

The target to sell half of the position

E projection – Up: E = B + (B - A) Down: E = B - (A - B) [Figure 9]

• Up: E projection adds the distance of the last upleg to the last
high.
• Down: E projection subtracts the distance of the last downleg
from the last low.

This is the target to sell all the positions.

PAGE 78 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Figure 9. E Projection
Figure 9. E Projection
Up: E = B + (B - A) E projection Down: E = B - (A - B) and lows) are used (Figure 12). When the market is about to
make a big move up, typically, those distances are multiplied by
E A
four to project higher targets.
C
= Figure 12.
Figure 12.Four timestimes
Four higher ishigher
the targetisinthe
a bigtarget
move up in a big move up

B
=
Up: 3E = B +3 (B - A) 3E projection Down: 3E = B -3 (A - B)

B A C
3E
= =
C
= =
2E B
A E
= =
E
The target to sell all the positions E
B = =
NT projection -- Up: NT = C + (C - A) Down: NT = C - (A - C) 2E
[Figure 10] = =
• Up: NT projection adds the distance between the last two lows 3E
A C
to the last low.
• Down: NT projection subtracts the distance between the last
The the
target Figure
Figure 13.13. Downward wave of Amazon
lastto sell all the positions Downward wave of Amazon.com
two highs from high.
Figure 10. NT Projection
Figure 10. NT projection
Up: NT = C + (C - A) NT projection Down: NT = C - (A - C)

A
NT

B =C
=
C =
= B

NT
A
Figure 11 is a daily chart of Wisesoft, which is listed on the Figure 13 is the downward wave of Amazon.com in the
Shenzhen market in China. In a short-term wave, a high was second half of 2018. The lower targets are multiplied by four
reached at N projection of 18.77, and a correction was reached in the initial range. When N wave extended to nine waves, the
at E projection of 13.72. And then, it was reached at V projection downtrend was over.
of 16.35 with a smaller upward N wave, and a correction was
reached at N projection of 12.7. Currently, you can see a larger Habitual Price Range (applied type)
downward N wave because it was below the previous low. There is also an applied type that uses past price range to
predict future price range (Figure 14). This is the idea that the
Figure 11.
Figure 11. Daily
Daily chartchart of Wisesoft
of Wisesoft price range from A to B and from C to D is the same.
Figure
Figure 14.14. Habitual
Habitual price range
price range(applied type) (applied type)

In addition to the four basic projection methods (N, V, E and

NT projection), it is also effective to use the middle-point of each Figure 15 is a daily chart of ELSWEDY ELECTRIC, which is
projection. listed on the Cairo market. Although V, N, and E projections
Moreover, to project higher-degree targets, using the four can be seen in various places, the left and right range excluding
basic projection methods, whole-number multiples of the intermediate waves are almost the same.
distances used above (e.g., distances between previous highs

IFTA.ORG PAGE 79
 IFTA JOURNAL 2021 EDITION

Figure
Figure 15.15. Daily
Daily chart
chart of of ELSWEDY
ELSWEDY ELECTRIC ELECTRIC
Reversal dates are the dates on which the market is projected
to “reverse” directions at relatively high probabilities. However,
the market does not always “reverse” on a Reversal date. In a
strongly trending market, the existing move sometimes simply
“accelerates” instead of “reverses” on a Reversal date. This
happens more often in a down-trending market than in an
up-trending market. Suppose there is a market that has been
moderately declining into a Reversal date. If it cannot reverse
direction during that time window, oftentimes it starts falling
sharply.
Figure 17 is a daily chart of CIMB Group, which is listed on the
Malaysian market. It shows that the market direction changes
when it reaches from a major low to a Basic number or a near
Time Principle Basic number.
The third subject is the time projection. One striking
characteristic of the Ichimoku theory is the degree of Figure
Figure 17.17. Daily
Daily chart chart
of CIMB of CIMB Group
Group

importance it places on the time factor. Hosoda taught, “It is not

that time merely passes as prices fluctuate in the market. Time
influences the market. The market is dictated by time.”
Ichimoku calculates “Reversal dates” in the following “Basic
number”-based projection and “Time parity”-based projection.
These two methods can be used separately or simultaneously.

a) “Basic number”-based projection

First, I explain “Basic number”-based projection. Following
(Table 1) are the “Basic numbers” to be used with period data
as default parameters. In addition to that, there are 83,97,101,
etc. Some Ichimoku researchers claim that they have found that
5,13,21 should be added as Basic numbers when dealing with
weekly data. b) “Time parity”-based projection
Second is “Time parity”-based projection. In this method,
Table 1. “Basic
Table 1. “Basic numbers”
numbers” to be used with to bedata
period used with period data Reversal dates are projected by adding the same time distance
Basic number Comments
9 ‐Useful for calling intermediate tops/bottoms
between two key dates in the past to the high or low on which
17 ‐Useful for calling intermediate tops/bottoms (9+9-1) the trend reversed, from which to project into the future. Figure
26 ‐First term, Useful in up markets (9+17) 18 illustrates this.
33 ‐Particularly useful in down markets (17+17-1)
42 ‐Very important in both up and down markets (17+26-1) Figure
Figure 1818. “Time
“Time parity”-based
parity”-based projection projection
51 ‐Second term (26+26-1)
65 ‐More useful in up markets than in down markets (33+33-1)
76 ‐Third term, More useful in up markets than in down markets (26+26+26-2)
129 ‐More useful in up markets than in down markets (65+65-1)
172 ‐More useful in up markets than in down markets (33+65+76-2)

Reversal dates are projected by adding “Basic numbers” to

the high or low on which the trend reversed. Figure 16 illustrates
this.
Figure
Figure 1616. “Basic
“Basic number”-based
number”-based projection projection

1. Add the time distance (the number of the days) between the
high C and the low D to the date of the low D into the future.
2. Add the time distance (the number of the days) between the
low B and the low D to the date of the low D into the future.
3. Add the time distance (the number of the days) between the
high A and the low D to the date of the low D into the future.
4. Add the time distance (the number of the days) between the
high A and the high C to the date of the low D into the future.

PAGE 80 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Basically Nine Patterns of “Time Parity” correction, the uptrend will continue until the coincidence point
There are basically nine patterns of taking Time parity of time and price is reached.
numbers. Of these, three are shown in the circles(Figure 19). The
left one means that the first I wave determines the V wave in the Figure 21.
Figure 21. Coincidence
Coincidence point ofpoint ofprice
time and time and price
next identical period. And the right one means that the N wave
determines the V wave in the next identical period.
Figure
Figure 1919. “Time
“Time parity”-based
parity”-based projection projection

Figure 22 is a daily chart of ORIENTAL WEAVERS as a sample

of the N projection. On the left side of the chart, the coincidence
point of the number of the time parity 54 and the N projection in
the N wave was the Reversal date for the upward wave. On the
Figure 20 is a daily chart of SIDI KERIR PETROCHEMICALS right side of the chart, the coincidence point of the Basic number
(Sidpec), which is listed on the Cairo market. Starting from a 65 and the N projection in the N wave was the Reversal date for
low in May 2017, one would notice that the Time parity-based the next upward wave.
projection worked well to call market turns. Also, one would
notice that the numbers are close to the Basic numbers. To be Figure 22.
Figure 22. Daily
Daily chart
chart of of Oriental
ORIENTAL WEAVERS Weavers

sure, separately run Basic number-based projection and Time

parity-based projection often converge.
Figure 20.
Figure 20. Daily
Daily chart
chart of of Sidi
SIDI KERIR Kerir Petrochemicals
PETROCHEMICALS

I hope that you all have understood the relationship between

the “wave structure,” “price level,” and “time” principles.

Forecasting Major World Indices

Coincidence Point of Time and Price Is the Dow Jones Industrial Average, Monthly
Important Reversal Date Figure 23 is a monthly chart of the Dow Jones Industrial
Finally, I’d like to summarize the main point. Determine Average. In July 2019, a high was reached at N projection of
important Reversal dates such as reversal or acceleration by 27,251 dollars in N wave from March 2009. After that, the stock
recognizing the coincidence point of the price target on the price dropped sharply; however, San’yaku Koten continued. And
vertical axis and the time on the horizontal axis. the conversion line has accelerated in October 2019; therefore,
For example, in the upper diagram in Figure 12, the the market became more bullish. As a result of responding to
coincidence point of the price target calculated using N changes in the conversion line, November 2019 was a significant
projection and the Basic number 76 from the low is the acceleration point, as it was the month with the time parity 12
important Reversal date. and the basic number 129.
In the lower diagram in Figure 12, the coincidence point of the And the 2V projection of 29,654 dollars from a low in 2009 was
price target calculated using 2E projection and the Time parity reached in February 2020. The February high was close to time
number 81 from the low is the important Reversal date. “81” is parity number 25 from the January 2016 low to the January 2018
the number of days in 29 and 53 V waves in the past N waves. high. It is a basic number 26 in the first place. This means that
It means that the time parity number 53 does not coincide the time and the price did coincide. After that, the stock price
with price projection. Therefore, one should judge that after dropped sharply. If the market can stay San’yaku Koten after the

IFTA.ORG PAGE 81
 IFTA JOURNAL 2021 EDITION

Figure 24. Monthly chart of the Shanghai composite

correction, one should judge that the uptrend will continue until Figure 24. Monthly chart of the Shanghai composite index
index
the coincidence point of time and price is reached again.
Basically, the points where the trend is likely to change are
the timing when the Lagging span touches the stock price. If
the Lagging span falls below the stock price, a bearish sign is
confirmed. However, as with the 2016 correction, if the Lagging
span jumps to support the stock price at that time (January
2014), the trend will be further strengthened.
Figure 23. Monthly chart of Dow Jones Industrial
Figure 23. Monthly chart of Dow Jones Industrial Average
Average

EGX30 Index, Weekly

Figure 25 is a weekly chart of the EGX30 Index. Currently,
it is a small-scale rebound from the low in December 2018. If
it exceeds the high in February 2019, N wave from the low in
December 2018 can be confirmed. And you can predict N(16457)
and E(18566) upward projections. In addition, it is important to
predict the middle-point (17511) of N and E projections. After the
short correction, if it exceeds the high in April 2018, a larger N
waves starting from the low in January 2016 can be confirmed.
If it exceeds the high in February 2020, it will be the ninth In that case, the price range of 12701 from the low in January
wave. The upside target is E projection of 30,232 dollars from 2016 to the high in April 2018 and the price range of 6338 from
the low in 2009. Also, there is V projection of 32,190 dollars in the high in April 2018 to the low in December 2018 will be an
the period after 2018. This is an important target that consistent important factor for the price projection.
with 3E projection of 32,097 dollars from the low in 2009. And However, if it falls below the low in December 2018, the
the most important timing on the time projection is from July medium-term N wave starting from the high in April 2018
to October in 2021, from January to April 2022, December 2022, can be confirmed. And you can predict N, V, and E downward
and June 2023. projections.
After 2020, what is important in the time projection is the
Shanghai Composite Index, Monthly time parity 34 for short-term wave and the time parity 66,120
Figure 24 is a monthly chart of the Shanghai composite for medium-term waves. In addition, the week coming with the
index. Look at the long-term N wave from a low in 2008. The basic number 76,83,97,101 from the low in December 2018 and
coincidence point of 81 months close to the Basic number 83 the basic number 129,172 from the high in April 2018 will be the
and the middle-point(VE:5199) of E and V in the N wave was the important reversal weeks.
reversal month for the upward wave.　 I have just written about the future assumptions so far;
Currently, it is moving in the downward wave from a high in however, what I want to say is, regardless of the size of N wave,
2015. In the short term, the Conversion line may go above the whether it be an uptrend or a downtrend, it will continue until
base line; however, it is not just San’yaku Koten. That is, the the coincidence point of time and price is reached.
downward wave can be predicted to continue. In that case, the
coincidence point of the time parity 81, 93, 97 and V projection Figure 25.
Figure 25. Weekly
Weekly chart ofchart of EGX30
the EGX30 Index Index
of 1,689 in the N wave from a high in 2015 is likely to be an
important reversal month for the downward wave. In addition
to focusing on the move from a high in 2015, the time parity 13,
32, 37, 44 may also be important as the time on the horizontal
axis. The downside target is V projection of 1,689 in the N
wave from a high in 2015 as well as N projection of 2,141 and V
projection of 1,592 in the N wave from a high in 2018.
If it exceeds a high in 2018 with San’yaku Koten, it is likely to
lead to a larger N wave starting from a low in 2013. In this case,
the month coming with the time parity 32, 37, 44, 68, 81, 93, 97
will be the important reversal month for the upward wave. And
it is considered that the upward wave will continue until the
N projection of 5,698 or the level of 4,909 obtained using the
habitual price range.

PAGE 82 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Nikkei Stock Average, Monthly General Disclaimer

Figure 26 is a monthly chart of the Nikkei Stock Average. This report has been produced by the author for informational
When you look left and right centering on a high in February purposes only. It does not constitute solicitation of the sale or
2007 and a low in October 2008, you will notice that the purchase of securities or other investments. The information
time parity numbers are major highs and lows. In this case, contained herein is derived from sources that that author
the months coming with the time parity 160–207 will be believes to be reliable, but the author does not guarantee the
the important reversal timing for the upward wave or the accuracy or completeness of said information. In some cases,
downward wave. such information may be incomplete or summarized. Prices,
And after a low in October 2008, what is important in the time numbers, and similar data contained herein include past results,
projection is the time parity 29,41–44,56,81 for medium-term estimates, and forecasts, all of which may differ from actual
waves. If the price projection (N,V,E, etc.) is realized at the time data. These prices, numbers, and similar data may also change
parity 29,41–44,56,81 for medium-term waves, this will be the without prior notification. This report does not guarantee
important reversal month. future performance, and the information contained herein
should, for whatever purpose, be used solely at the discretion
Figure
Figure 26.26. Monthly
Monthly chart ofchart ofStock
the Nikkei Nikkei stock average
Average
and responsibility of the reader. It is prohibited for this report to
be provided or redistributed to any third person, or modified or
changed in any way.

IFTA.ORG PAGE 83
 IFTA JOURNAL 2021 EDITION

Stock Trends and Trend-Based Trading c/o NAAIM Founders Award

Strategies—Backed by Large-Scale Back-Testing 6732 W. Coal Mine Avenue #446

Littleton, CO 80123

Implemented by Automated Software System Phone: +1 (888) 261-0787

Web: https://www.naaim.org
By Kevin Luo

Figure 1. Trend chart sample. Source: Archer Daniels

Introduction Midland Co. (ADM)
The term “stock trends” may evoke many bright and
dark images: rising Apple stocks, comeback of the NASDAQ
Composite Index in 2014, meltdown of the financial sector in
2008, and more. What are stock trends? Do all investors who
use stock trend analysis in their investment decision-making
process have a clear view of their stock trends? It can always be
argued if stock trends should be vaguely or precisely defined.
It cannot be argued that vagueness creates gray areas. Gray
areas grow uncertainties that can lead to serious risk problems.
This paper describes and discusses a research project in which
stock trends were quantitatively defined with a single factor, The close prices in green belong to the uptrends and the
isolated, analyzed, and then summarized. The method is easy to close prices in red belong to the downtrends. Each trend has its
approach. Its results are precise and black/white. The paper also beginning and ending prices, which are jointly used by linking
demonstrates applications of the trend analysis with introduction trends. The numbers (in 100%) on the chart represent trend
of two trading strategies formulated based on a trend technique. sizes of the respective trends. Trend size is the vertical distance
between beginning and ending levels of a trend divided by the
Trend Definition and Description beginning or ending price. Using the term “trend size” is an
Edwards and Magee (2007) have developed some exclusive attempt to differentiate the concept from existing terms such
frameworks for trend analysis based on the Dow Theory. as “trend length,” which factors in time frame. In terms of
According to Edwards and Magee, there are two decisive equations, trend sizes are:
factors in trend determination. They are price movement and 1. Uptrend Size = (Ending Price – Beginning Price) / Beginning
time frame of such price movement. In terms of “primary Price
trend,” it is “appreciation or depreciation in value of more 2. Downtrend Size = (Beginning Price – Ending Price) / Ending
than 20%” and “last for a year or more” (Technical Analysis of Price
Stock Trends, Edwards and Magee, 2007, P15). This approach Please refer to Appendix One of the Appendices Section for
provides some ultimate leads for practical applications of details regarding the difference of uptrend and downtrend
stock trend analysis. Among the two trend decisive factors, calculations.
price movement is obviously independent and time frame
is dependent. Time frame is an important factor; however, Trend Isolation and Trend
it sometimes creates confusion. Questions would definitely Characteristics Analysis
be raised when a market declines 20% and then climbs right To prove that the trend analysis with only price factor
back in less than one year. Do we classify the price movement still works, the project was designed to isolate stock trends
as a stock trend? Can money be made or lost during the price from a large number of stocks and then summarize trend
movement just like in a “qualified” primary trend? Would characteristics, which are in turn to be used in the trading
trend analysis still work without the time frame? A research strategy development. If the developed strategies work, the
project was conducted to isolate and analyze stock trends trend analysis works. As the larger scale leads to increased
without the time frame factor. In the project, two types of precision and reliability, 1,816 random stocks were selected for
trends were defined. Uptrend is recognized in case prices move the research. These stocks were all listed on NYSE and NASDAQ.
upward more than 20% and downtrend is recognized in case Standard daily price data (open, high, low, and close) was used.
prices move downward for more than 20%. The definitions Data range covered an approximately 10-year period of time
are practically identical to the condition of “appreciation from January 1, 2005 to September 1, 2014. Tasks of the project
or depreciation in value of more than 20%” by Edwards and such as trend isolation, characteristics summarization, back-
Magee. A sample trend chart is exhibited in Figure 1. testing, and graph generations were automatically implemented
by a custom-designed software system (the software).

PAGE 84 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

In the trend isolation process, the software separated 88,353 sections. Among interpretations of this finding, one stands out
stock trends from 1,816 stocks. Figure 2 shows the average trend certain and clear. There must be good trading opportunities
sizes by different price range groups. On average, the uptrend for the remaining section of the typical trend. The remaining
size measured 48.29% and the downtrend size measured 45.37%. section ends when the trend reverses. For the typical uptrend, it
covers 28.9% of the uptrend (after the confirmation section) and
Figure 2. Average trend size (%) by stock price group 20% of the following downtrend (or the downtrend confirmation
section). Trading strategies may be developed targeting the
28.9% in the uptrend or the
entire remaining section. In this research project, two
sample trading strategies were formulated targeting the
entire remaining section. The remaining section is labelled
as “trading range” in this paper for convenience. A sample of
uptrend trading ranges (gray-shaded sections) is displayed in
Figure 5. Next, the paper details the strategy development and
back-testing of two trading strategies. The back-testing of the
strategies determines effectiveness of the trend analysis.
The study found no evidence that the cheaper stocks produce
the larger trends; however, as shown in Figure 3, the cheaper Figure 5. Uptrend trading range sample. Source: Archer
Daniels Midland Co. (ADM)
stocks did generate larger trend counts than the expensive
stocks. The stocks priced under $5 produced approximately
three times more uptrends and downtrends than stocks priced
over $40 on average. This probably means that the cheaper
stocks tend to move up and down more frequently. Therefore,
it is reasonable to conclude that the average uptrend size of
48.29% and downtrend size of 45.37% represent the stocks in all
price ranges.
Figure 3. Average trend count by stock price group

Trading Range Strategy

Strategy
In this paper, trading strategy is defined as a set of methods
or techniques that can be independently and repeatedly used in
trading operations.
Trading Range Strategy was formulated based on the trend
characteristics obtained from the trend analysis. The strategy
In practice, a trend is not recognized until it is confirmed. aims to maximize per-trade profit with a few simple rules. As
The confirmation requires a stock to move a minimum of 20% stated, the average uptrend is measured at 48.29%, which is
in a direction. The confirmationportion of the trend is called more than twice larger than a confirmation section. One of the
“confirmation section” in this paper. In Figure 4, the sample strategy ideas was to take and hold a position for the whole
uptrend confirmation sections are covered by the green shades trading range. Figure 6 illustrates a sample trade of the strategy.
and the downtrend confirmation sections are in the red shades. The blue line connects buy and sell points. It is important to note
that this strategy only allows trades in the trend directions.
Figure 4. Trend confirmation section sample. Source: Only long trading is discussed in this paper.
Archer Daniels Midland Co. (ADM)
Figure 6. Trading Range Strategy trade sample. Source:
Archer Daniels Midland Co. (ADM)

One of the important findings in the research is that stocks

tend to move substantially further beyond confirmation

IFTA.ORG PAGE 85
 IFTA JOURNAL 2021 EDITION

Mathematically, after taking 40% out of both confirmation A set of test trade samples are graphically presented in Figure
sections, there is still 8.29% left. This 8.29% is the theoretical 8. The blue lines connect buy and sell points of the trades.
per-trade profit for the strategy; however, assuming 8.29% as
a figure close to the practical profit is very risky because the Figure 8. Trading Range Strategy back-testing sample
output. Source: MicroStrategy Inc. (MSTR)
trading is in the same directions with the stock trends. In many
cases, price actions such as open-gaps would easily wipe out
all of the profits. In Figure 7, the sample shows the impact of an
open-gap at the entry. Without the open-gap, the entry would
have been recorded at the projected entry price (orange-line
level). With the open-gap, it cost around $1.50 more to get in
the long position (green line level). According to the statistics,
32.43% of trades were affected by open-gaps. Sometimes the
open-gaps are very large. Skipping all open-gaps that cause
higher than projected entries is a practical alternative; however,
such attempts often end up losing good opportunities at the
same time. After implementation of the test tasks, the software
produced some summary statistics for evaluation. The
Figure 7. Open-gap impact on entry price sample. Source: computation approach and equations are discussed in Appendix
Akamai Technologies Inc. (AKAM)
Two of the Appendices section.

Back-Testing Outputs
The software generates a set of performance statistics,
including cumulative P&L, drawdown (graphically presented
in charts of the Figures), average annual return, average per-
trade return, Sharpe ratio, and max drawdown (in footnotes of
the figures). In Figure 9, the summary shows that the strategy
generated a 16.90% average annual return for the Top 500
Portfolio. The per-trade returns averaged at 3.95%, which is
much smaller than the theoretical figure of 8.29%. The Sharpe
It is possible but unlikely for a simple strategy like Trading Ratio was computed based on annual returns. The figure was
Range Strategy to work well on all stocks and all the time. It calculated as 0.54. On the risk side, the max drawdown was
is necessary to select some suitable stocks from the random recorded as 48.00%. Surge of the cumulative return in the end
stocks for the back-testing. In theory, the larger the trend is, of chart was resulted from liquidating all positions described in
the greater the profit is. The suitable stocks for this strategy Rule 2 of the back-testing. The software books profit/loss of the
are considered as those with the larger trends. To confirm the test trades at the exits.
theory, the research selected the top 500, top 200, and top 50
stocks at the top of a trend size sorted list. The back-testing Figure 9. Back-testing performance (500 stocks);
Trading Range Strategy back-testing.
generated summary statistics separately for the portfolios.
These trading portfolios are labelled as Top-500 Portfolio,
Top-200 Portfolio, and Top-50 Portfolio. The stock prices in the
portfolios average $21.81, $19.94, and $13.83 for the Top 500
Portfolio, Top 200 Portfolio, and Top 50 Portfolio, respectively.
The average trend sizes are 51.67%, 54.13%, and 56.12% for the
respective portfolios.

Back-Testing
Back-testing is one of the most effective approaches for new
strategy evaluations. In the back-testing of this project, the
software tests each of the stocks individually by generating For the Top 200 Portfolio, the average annual P&L arrived
trades according to the trading rules and then summarizes the at 25.90% which, as expected, is approximately 50% higher
test outputs by the trading portfolios. No costs or volumes are than the return from the Top 500 Portfolio (Figure 10). The
factored in. average per-trade return is also higher by approximately 35%
There are a couple of rules for the test: at 5.32%. The Sharpe Ratio and max drawdown are and 44.50%,
1. Enter a long trade at the first available price after each of the respectively.
uptrends is confirmed.
2. A trade exits at the first available price after the uptrend
reverses or at the end of the test. Close price is used for
liquidating a position at the end of the test.

PAGE 86 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Figure 10. Back-testing performance (200 stocks);

Trading Range Strategy back-testing Using a stock index as the benchmark for the performance
comparison may cause concerns in this case because this
strategy involves frequent trading; however, such comparison
does make some sense at this level of evaluation. This research
puts performance statistics of the back-testing and two stock
composite indices together for the comparison and evaluation.
The two stock composite indices are NASDAQ (^IXIC) and NYSE
(^NYA) of which the stocks in the back-testing belonged to.
In comparison of the cumulative returns, it appears that the
portfolios outperformed the indices most of the time (Figure
13). The slopes of the cumulative return curves for the portfolios
In Figure 11, the test performance for the Top 50 Portfolio is could have been adjusted steeper upward for 2012–2014 if the
presented. The strategy worked better for this portfolio than unrealized profits (the pop-up amounts in the end of the curves)
the Top 200 Portfolio by 42%, as the average annual return were accounted for.
reached 37.00%. The per-trade return of 6.11% is also greater
than the 5.32% of the Top 200 Portfolio. The Sharpe Ratio is 0.77 Figure 13. Test performance comparison; Trading Range
Strategy vs. indices
and the max drawdown is 47.10%.
Figure 11. Back-testing performance (50 stocks); Trading
Range Strategy back-testing

Observations from Figure 13 (or Figures 9–11) also indicate

that the strategy took some large losses. The drawdown
readings reached as low as 48% for the Top 500 Portfolio. Those
Evaluation drawdown figures were primarily driven by the substantial
Evaluation of the Trading Range Strategy focuses on the back- declines of overall markets during 2008–2009. Under such
testing performance, including returns and risks. Comparing market conditions, the uptrends shrunk and the larger open
returns among thethree portfolios can determine if there is gaps at the exits pushed the returns further in the red. There is
a tendency for stocks with the larger trends to outperform no easy fix for this issue unless adding managerial techniques
other stocks. As shown in Figure 12, the Top 200 Portfolio such as long/short strategy or optimizing the trading model.
outperformed the Top 500 Portfolio and the Top 50 Portfolio On the bright side, the losses were not as bad as the indices
topped the Top 200 Portfolio by visible amounts consistently suffered. The indices have carried up to 74% and 85% drawdowns
throughout the years. It confirms that there is a positive throughout the period of time. This resulted from the trading on
relationship between the stock trend size and the performance the uptrends only. The stocks have stayed on the sideline during
for the strategy. Knowing this relationship is helpful in future the larger-than-average downtrends.
optimization and trading operations.
MACD Strategy
Figure 12. Test returns by portfolio and year; Trading Back-testing results of Trading Range Strategy have proved
Range Strategy back-testing
that there are good opportunities for trading range in the
trend analysis. Now the question is to what extent this trend
technique can be applied. Here is another example strategy
that connects trading range with one of the popular technical
indicators, Moving Average Convergence/Divergence (MACD).
This strategy is called MACD Strategy because MACD is the main
base of the strategy.
Created by Gerald Appel, MACD is considered a simple and
effective momentum indicator. MACD signal line crossover is
a method showing bullishness and bearishness of the prices. A
bullish crossover occurs when the MACD line rises and crosses
the signal line from below. A bearish crossover happens when
the MACD line falls and crosses the signal line from above.
According to statistics obtained from the trend analysis of

IFTA.ORG PAGE 87
 IFTA JOURNAL 2021 EDITION

Figure 15. MACD strategy short trades. Source: Allegheny

the 1,816 random stocks, the average price changes (calculated Technologies Inc. (ATI)
as Close today – Close Prior in percent) are 2.18% and -1.92%
for the bullish and bearish crossover days, respectively. These
figures are considered significant when compared to 0.44% and
-0.55% of price changes for the typical uptrend and downtrend
day. It appears that MACD timely signals bullish/bearish price
momentums. Can the signals be used as trading signals for
long/short stocks? A back- testing was conducted to find the
answer to the question. Test rules for long trades were buying at
bullish crossover and selling at the following bearish crossover.
The rules for short trades were shorting at bearish crossover
and covering at the following bullish crossover. Close prices
of the signal days were used as entry and exit prices. A timing Figure 16. MACD strategy short trades with trading
sample is exhibited in Figure 14. The up arrow points to a bullish range. Source: Allegheny Technologies Inc. (ATI)
crossover and the down arrow points to a bearish crossover.
Figure 14. Sample MACD signal line crossover timing

The summary of the second test shows that trading range

improved the long trading return by 9.97% and the short trading
return by 43.71% for the 1,816 random stocks. For the Top 200
Back-testing of MACD signal line crossover was implemented Portfolio, the average annual returns were recorded as 16.82%
by the software that generated long or short trades for each for the long trades and -8.34% for the short trades. The test
crossover of the stocks. The average annual return for the long results are strong enough to conclude that the trend technique
trades is 1.45% but the short trades lost (15.94%) annually on can be effectively incorporated with external strategies for
average. A finding of this research suggests that there are a performance improvement.
large number of short trades in the uptrends and long trades in
the downtrends. MACD is a good momentum indicator as the Conclusion
statistics suggest; however, the momentum signals generated Success in testing the strategies suggests that stock trends
by MACD are probably too frequent for a normal low-frequency defined with price factor alone worked effectively. The trend
trading strategy. characteristics analysis offers valuable statistics for many
When trading against the trends too often over time, the areas of the investment decision-making process. It is easy
profits are expected to be poor. Trading range is supposed to be to approach and does not require a great deal of effort to
capable of reducing some of the unfavorable trades to improve understand. Outputs of the trend analysis are precise. There is
the performance. Another back- test was conducted with the no gray area of confusion. From a strategy design standpoint,
trading range technique factored in. A couple of new rules were the trend analysis offers great flexibilities. Stock trends can
added. be isolated into major, intermediate, minor, or custom-defined
1. MACD long trades are only allowed in uptrend trading ranges trends to meet user needs. There is huge potential for the
while short trades are only allowed in downtrend trading techniques to grow in new strategy development and existing
ranges. strategy improvement.
2. At the end of the trading ranges, all holding positions are Trading Range Strategy is one of the simplest but most
liquidated at the close price of the day. effective trading strategies. It is straightforward and easy to
follow. There is large room for optimization of the strategy.
Impacts of the trading range technique can be easily observed Through the introduction of MACD Strategy, this paper has
by comparing sample test trades shown in Figure 15 and Figure 16. shown how effective the trend techniques can be in terms
Figure 15 indicates the short trades generated without trading of incorporating with an external strategy. Results from
range for Allegheny Technologies Inc. (ATI). In Figure 16, the the separated projects have confirmed that trading range
chart shows the short trades for ATI with trading range added. technique also worked well with some other technical
The trade count reduction is significant and the performance indicators such as RSI.
improvement from -10.71% to 7.30% is impressive.

PAGE 88 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Appendices For the purpose of comparison, test results are also calculated
conventionally. It utilizes equal-weighted approach, in which
Appendix A $1,000 is assigned to every trade regardless of share prices. The
Conventional calculation of downside percentage works for related equations are listed below.
the majority of situations, but not the downtrend calculation
in this research. The easiest way to describe the problems is to 1. Number of Shares (Trade) = $1,000 / Entry price
show an example (Figure A1).
2. P&L (Trade) = Exit Price * Number of Shares - $1,000
Figure A1. Downtrend calculation method comparison
Source: Archer Daniels Midland Co. (ADM)
3. Daily P&L (Portfolio) = P&Ls (All Daily Completed Trades) /
Portfolio Value (100%)

4. Cumulative P&L (Portfolio) = (Cumulative Daily P&L (Prior to A

Given Day) + Daily P&L (A Given Day)) / Portfolio Value (100%)

5. Average Annual Return (Portfolio) = End-Test Cumulative P&L /

Years (100%)

Because there were no cost or volume factors in the

In this case, the stock starts to move lower after reaching performance computation, the process was simplified by using
the uptrend (green bars) high. By the conventional method, a fix value as “Portfolio Value.” “Portfolio Value” is computed
which is expressed in Equation 1 at the end of this paragraph, based on stock counts in a portfolio. Trading a portfolio of
the uptrend would not reverse until the stock reaches stocks requires less capital to operate than holding the same
approximately $36 (red line level in chart). That level is stock portfolio. It is unreasonable to use full counts of stocks
significantly lower than where the uptrend begins. It is for “Portfolio Value” computation in this case. The chart in
reasonable to assume that this is a major concern for the Figure A2 shows the stock holding counts for the back-testing of
majority of the trend users. To fix the problem, this research Trading Range Strategy.
employed an alternative method which utilizes an easy equation
(Equation 2). With this equation, an uptrend reversal level Figure A2. Number of holdings—Top 500 stocks; Fixed
portfolio value calculation
never falls below its beginning level anymore (blue line level
in Figure A1). More importantly, the alternative method makes
the trend analysis more meaningful. Under the convention, the
maximum size for a downtrend is 100%. It is not comparable
and combinable with an uptrend because the uptrend can
be measured up to infinity percent. Without the alternative
method, simple strategies such as Trading Range Strategy
would have been more difficult to formulate. In this research, all
statistics tied to downtrend measures were computed based on
the alternative method. The less it holds, the smaller the capital needed to operate.
For this research project, the 95th percentiles were used for
1. Downtrend-Size (Convention) = “Portfolio Value” calculation. In this particular example, the
(Ending Price – Beginning Price) / Beginning Price (100%) 95th percentile (red line level in Figure A2) is 420 counts.
“Portfolio Value” is therefore calculated as $420,000
2. Downtrend-Size (Alternative) = ($1,000*420). $1,000 is the amount assigned to each trade.
(Beginning Price – Ending Price) / Ending Price (100%) The figure is smaller than $500,000 ($1,000*500) for a holding
portfolio.
Appendix A
Back-testing returns are expressed in two different ways. References
The first one is average per-trade method, which is very Robert D. Edwards and John Magee (2007), Technical Analysis of Stock
straightforward. The equations used in the calculations are Trends, 9th edition
exhibited below. John J. Murphy (1999), Technical Analysis of the Financial Markets
Christian L. Dunis, Jason Laws, and Patrick Naim (2003), Applied
1. Per-trade P&L = Quantitative Methods for Trading and Investment
(Exit Price – Entry Price) / Entry Price (100%)

2. Average per-trade P&L (trading portfolio) =

sum of all per-trade P&Ls (all stocks) / Number of Trades (all
stocks)

IFTA.ORG PAGE 89
 IFTA JOURNAL 2021 EDITION

Lewis A. Glenn, Ph.D.

Simple and Effective Market Timing Chief Scientific Officer
Creative Solutions Partners Associates
With Tactical Asset Allocation 702 Park Hill Rd.
Danville, CA 94526 US
[email protected]
By Lewis A. Glenn, Ph.D.1 Phone: 925-820-5158

Abstract by more than a factor of five and her holdings would be down
A simple market timing algorithm is examined that switches to less than half the initial investment. It would take five more
from an exchange traded fund representing years for this investor to recoup the original investment, not
U.S. equities to one holding treasury long bonds every month including the interest she would have made had she remained in
on the last day, the switch being made to whichever ETF has cash. Her willpower would once again be tested as she watched
the greatest ratio of current adjusted closing price to adjusted the collapse of the financial markets in 2008, by the end of
closing price µ months earlier. The parameter µ is determined so which the value of the initial investment would be once again
as to maximize total return and minimize the total number of halved.
trades, however the results are relatively insensitive to µ over a It is this scenario that tactical asset allocation seeks to
fairly wide range. The performance of this scheme is compared mitigate. The main idea is simply to diversify portfolio assets
to that of an Ivy 5 portfolio consisting initially of equal dollar and employ a market timing solution. But what kind of solution?
amounts of ETFs in U.S. equities, foreign large blend, 7–10 year The literature is replete with different approaches. Our goal here
treasuries, real estate, and commodities. As with the paired is not to review the many ideas that have been suggested but
switching approach, each ETF is purchased only once a month, rather to focus on two very simple schemes.
on the last day, in this case only if its adjusted closing price An especially popular method was described by Faber3 and
exceeds the 10-month simple moving average (SMA). Otherwise, then further elaborated by Faber and Richardson4 as the Ivy 5
that portion of the portfolio is invested in a cash surrogate. portfolio. The basic idea is to create a portfolio of five sectors
Comparison is made over the 10-year period ending on consisting of the S&P 500 index, the MSCI EAFE index, U.S. 10-
12/31/13. It is shown that the average annual return of the year government bonds, a real estate index, and a commodity
paired switching algorithm exceeds 30% in this period, which index. The initial study involved non-tradable indices over a
is three times greater than that of the Ivy 5. Moreover, only 35-year period. The trading rule was simply to buy each index
45 trades were required for the paired switching approach, when the monthly price exceeded the 10-month simple moving
whereas the Ivy 5 required 70 in the same period. The maximum average (SMA) and to sell and move to cash otherwise. All entry
draw down was 14.6% for the Ivy 5 and 18.8% for paired and exit prices were on the day of the signal at the market close
switching. and the model was only updated once a month on the last day.
Price fluctuations during the rest of the month were ignored.
Introduction Moving average methods are very useful for discerning trends
The so-called weak form of efficient-market theory (EMT) and indeed this author has suggested several market timing
holds that future stock prices cannot be predicted on the basis algorithms based on them.5,6,7 All moving average algorithms,
of past stock prices2. Many advocates of this theory believe however, exhibit latency, namely, they introduce some lag in the
that the best strategy for the long-term investor is to “buy calculation so that the MA lags the original signal. This means
the market,” by which they generally mean to invest in an that the user will implicitly be late in leaving a declining market
index fund that represents all, or a significant segment of, the and, perhaps even worse, late in entering a rising one. The lag
equity market. The claim is that past performance has shown time increases directly with the length of the moving average.
that in the long run buying and holding this class of asset will The paired switching method to be described here does not
outperform any active management scheme. depend on moving averages. In general, paired switching refers
One problem with this approach is the definition of the term to investing in one of a pair of negatively correlated assets and
“long run.” To paraphrase the famous British economist John periodic switching of the position on the basis of the relative
Maynard Keynes, “…in the long run we’re all dead.” In fact, many performance of the two. Maewal and Scalaton8 proposed an
investors would consider a 5–15 year investment as long-term, especially simple version that looked at the performance over
whereas some would choose a period as long as 30 years and a prior 13-week period and purchased the asset that had the
others as short as six months. A brief look at the large cap higher return over that period. That position was then held for
growth ETF (QQQ) that tracks the NASDAQ 100 (which consists 13 weeks at which point the cycle was repeated. They looked at
of the largest non-financial securities listed on the Nasdaq Stock several different ETFs that served as surrogates for equities and
Market) exhibits what can happen over a long run. An investor bonds, including SPY (for the S&P 500) and TLT (for treasuries
purchasing this fund at its inception in March 1999 would have with maturities of 20 or more years).
seen her investment more than double in the first year. Two and Cohn9 devised a similar scheme but employed it on a daily
a half years later, by September 2002, this would have dropped basis. Instead of a look back period of 13 weeks (65 trading days),

PAGE 90 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Figure 1. Cumulative annualized growth rate for the

he found that 84 trading days produced the best results. pair switching algorithm for the 10-year period ending
In what follows we first generalize the Maewal-Scalaton 12/31/13 as a function of look back period. Also shown
procedure by switching between SPY and TLT in the same way as on the right ordinate axis is the total number of trades
with the Ivy 5, namely once a month on the last day, the switch made for each value of µ .
being made to whichever ETF has the greatest ratio of current
adjusted closing price to adjusted closing price µ months earlier.
The parameter µ is determined so as to maximize total return
and minimize the total number of trades. We then compare
the results to those obtained with the Ivy 5 portfolio over a
recent 10-year period and exhibit the results in a novel way
that allows an investor to immediately visualize end-of-period
performance based on date of first entry (purchase) with use of
the algorithm.

The Paired Switching Strategy

First, a few definitions are in order. Let the value of the
paired switching portfolio on day n be v(n), 1 ≤ n ≤ N and the
normalized value be V (n) = v(n) / C, where C is the cost basis. For
the 10-year period ending 12/31/13, N = 2517 trading days. Also,
let t(n) be the number of trades, where either a buy or a sell is
considered a trade.
For a buy and hold (BAH) strategy with an individual ETF, the
normalized value on day n is simply V ETF (n) = PETF (n) / PETF (1), Also of interest is the maximum draw down, defined as
where PETF (n) is the adjusted closing price on day n ≥ 1.
Then, the paired switching rule is simply:

Move to TLT on day n if and only if For the cases shown in Figure 1, and expressed as a percentage,
V TLT (n) / V TLT (n − µˆ) > VSPY (n) / VSPY (n − µˆ) Equation 1 δmax ( N ) was 18.8% uniformly over the range 2 ≤ µ ≤ 5 months;
this value first occurred on 4/30/09.
where µˆ is the daily equivalent of µ. Otherwise, move to SPY.
On all other days the previous position is maintained. Note The Ivy 5 Strategy
that daily price data on both SPY and TLT are available since In applying the Ivy 5 strategy we utilize the ETFs suggested
7/30/2002. by Faber and Richardson (FR)4 with one minor exception.
There were no broad sector commodity ETFs available on
Nominal trading costs are included on all transactions, but 1/1/2004, the beginning of our 10-year study period. So, we have
these costs are generally negligible with electronic trading and substituted the PIMCO Commodity Real Ret Strat Instl mutual
the high volumes extant with these ETFs. However, slippage fund (PCRIX) for the commodity ETF recommended by FR (DBC).
is not included, and this might be significant with increasing The modified Ivy 5 trading rule is then:
trading frequency. Purchase equal amounts of the following ETFs/funds on
To find the best value of µ to use, we first look to maximize day one if and only if the adjusted closing price exceeds the
total return, V ( N ) − 1, or equivalently the cumulative 10-month simple moving average (SMA): SPY (S&P 500), EFA
annualized growth rate (CAGR) over the entire period, χ( N ), (MSCI EAFE index), IEF (7–10 year treasuries), IYR (real estate),
where χ is defined as: PCRIX (commodities). Otherwise the amounts meant for
purchase are instead left in cash. Then, at the last day of this and
χ(n) = {[exp((log(V ) / (n / 250))] −1}×100 Equation 2 each following month, repeat the procedure. Price fluctuations
during the rest of the month(s) are ignored. Cash positions earn
(expressed as a percentage) and we have taken a trading year interest using IRX (CBOE index that measures the discount rate
to consist of 250 trading days. Note that since N = 2517 here, of the most recently auctioned 13-week U.S. Treasury Bill) as
Equation 2 evaluated at n = N will slightly underestimate the a surrogate. Again, nominal trading costs are included on all
true CAGR in the period of interest. transactions.
Figure 1 depicts χps ( N ) , as a function of the look back period,
µ. Also shown is the number of trades made. It is observed that Comparison of Paired Switching With the Ivy 5
maximum CAGR approaches 15% at µ = 3 months and that the Before proceeding, it will be convenient to slightly modify the
performance is relatively insensitive to µ between three and nomenclature. We will redefine the normalized portfolio value
five months. The number of trades decreases inversely with on day n > k as nkV where k ≥ 1 represents the day of purchase.
increasing µ in this period. Figure 2 then compares n1V for the paired switching algorithm
with that of the Ivy 5 portfolio. Also shown for reference are the
buy-and-hold (BAH) values of SPY, TLT, and BAH5 (the portfolio

IFTA.ORG PAGE 91
 IFTA JOURNAL 2021 EDITION

Figure 3. Total return on day N, the end day of the

consisting of initially equal dollar amounts of SPY, EFA, IEF, IYR, 10-year period beginning on 1/2/2004 and ending on
and PCRIX and in which no further trades are made). 12/31/13 for entry date corresponding to day k,
1 ≤ k ≤ N = 2517.
Figure 2. Portfolio value on day n,1 < n ≤ N = 2517, over the
10-year period n, beginning on 1/2/2004 and ending on
12/31/13

The open and closed circles represent end-of-month entry

dates, corresponding to the allowable entry dates with the
paired switching and Ivy 5 rules discussed above; purchases
made at other dates are assumed to be invested in cash, without
It is observed that N1V, the value on 12/31/13, for paired interest, until the next end-of-month, at which point the
switching is roughly double that delivered by the Ivy 5 portfolio, appropriate purchase is made. It can be seen that the paired
although, for the first half of the 10-year period, n1V for the switching scheme produces higher total return at the end of
former lags that of the latter by as much as 20%. The maximum the windowed period than the Ivy 5 independent of the entry
draw down over the entire period for paired switching was date. And, although an investor purchasing SPY at the end of
18.8% as against 14.6% for the Ivy 5. By contrast, N1δmax was February 2009 and holding through 12/31/13 would have done
55.2% for SPY, 26.6% for TLT, and 48.4% for BAH5. almost as well as with paired switching, it is highly unlikely that
Figure 2 compared results for every date in the period given he would have been able to identify the bottom of the market
a single entry (purchase) date, namely 1/2/04, the first day of that produced this result. Figure 4 shows the CAGR in the same
the period. More computationally intensive, but also of more context, and with similar modified nomenclature, as in Figure
interest, is what an investor might gain or lose at the end of 3. Over the entire set of entry dates the minimum, mean, and
the period for an investment made on any arbitrary date in the maximum values for paired switching are 14.4%, 20.4%, and
period. Figures 3–6 address this question. Figure 3 compares 31.9%, respectively. Note that the CAGR values after 12/31/12
the total return as a function of entry date, NkV −1, 1 ≤ k ≤ N , are suppressed for clarity because the results are highly
for paired switching with the Ivy 5 portfolio and with the three volatile and not meaningful for n / 250 ≪1 in Equation 2. The
references in Figure3. comparable CAGR minimum, mean, and maximum values for
the Ivy 5 are 1.4%, 6.6%, and 9.8%, respectively. Figure 5 shows
the maximum draw down for each of the cases in Figures 2–4
and exhibits the advantage of both paired switching and Ivy 5
schemes over buy-and-hold approaches.

PAGE 92 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Figure 4. CAGR on day N , the end day of the 10-year Figure 6. Number of trades on day N , the end day of the
period beginning on 1/2/2004 and ending on 12/31/13 for 10-year period beginning on 1/2/2004 and ending on
entry date corresponding to day k, 1 ≤ k ≤ N = 2517 12/31/13 for entry date corresponding to day k, 1 ≤ k ≤ N
= 2517

Figure 5. Maximum draw down on day N, the end day of

the 10-year period beginning on 1/2/2004 and ending on
12/31/13 for entry date corresponding to day k, 1 ≤ k ≤ N Discussion
= 2517 The results presented in the previous section would appear to
show that the paired switching scheme has the advantage over
the Ivy 5 unless draw down is the main concern, in which case
the significantly increased return of the former would need to be
balanced against its modest increase in risk. One question that
arises, however, is the significance of these results in predicting
future performance. For one thing, most of the ETFs involved in
both schemes have been in existence only a short time so only a
rather narrow 10-year period was considered. As mentioned ear-
lier, this problem was dealt with by Faber3 for the Ivy 5 by testing
the algorithm with non-trade-able indices over a 35-year period.
In their analysis of paired switching, Maewal and Scalaton8
employed the statistical procedure devised by Henriksson and
Merton10 to evaluate the significance of their results. They also
suggested a similarprocedure to that employed by Faber for the
Ivy 5, in this case the use of surrogate mutual funds—VFINX for
the S&P 500 and VUSTX for the treasury long bond. The former
has been in existence since 1980 and latter since 1986. Figures
7–11 show the results of our applying the paired switching algo-
rithm described in “The Paired Switching Strategy” with these
With neither timing scheme did the maximum draw down two mutual funds over a 25-year period beginning on 1/4/1988
exceed 20%, independent of entry date but the Ivy 5 had smaller and terminating on 12/31/13. In this case, the number of trading
drawdowns for most entry dates. Finally, Figure 6 depicts the days, N = 6966.
number of trades employed (each switch counts for two trades
with paired switching) for the Ivy 5 and paired switching
schemes. It is clear that the latter has the advantage regardless
of entry date. As mentioned earlier, even though the explicit
cost of electronic trading is now generally negligible, the cost
of slippage (the difference between the adjusted closing price
and the price actually obtained in a transaction) is not easily
accounted for. Slippage is not a deterministic process and, while
bounds may be established using Monte Carlo methods, this is
beyond the scope of the present study.

IFTA.ORG PAGE 93
 IFTA JOURNAL 2021 EDITION

Figure 7. Portfolio value on day 1 < n ≤ N, over the 25-year Figure 8. Total return on day N, the end day of the
period n, beginning on 1/4/1988 and ending on 12/31/13 25-year period beginning on 1/4/1988 and ending on
12/31/13 for entry date corresponding to day k, 1 ≤ k ≤ N
= 6966

Figure 9. CAGR on day N, the end day of the 25-year

period beginning on 1/2/1988 and ending on 12/31/13 for
Figure 7 shows that the portfolio value at the end of the period entry date corresponding to day k, 1 ≤ k ≤ N = 6966
with paired switching is double that obtained by buying and
holding the S&P 500 index surrogate and, as will be seen below,
this result was obtained with significantly less risk. In deriving
this result the trading rule Equation 1 was applied even though,
strictly speaking, extant short-term trading restrictions would
have prohibited its use. Moreover, when trading mutual funds,
the daily price is obtained only after the market closes so that
Equation 1 would need to be modified such that an indicated
trade at month’s end would need to be made at the following
day’s close, which would also avoid any slippage.
As before, Figure 7 compares results for every date in
the period given a single entry (purchase) date, in this case
1/4/1988, the first day of the period. Figures 8–11, by contrast,
exhibit the results at the end of the period for an investment
made on any arbitrary date in the period. Figure 8 shows
the total return as a function of entry date, NkV −1, 1 ≤ k ≤ N,
for paired switching and with buy-and-hold of VFINX and
VUSTX. The value of almost 2400% return over the 25-year
period amounts to almost 95% average annual return, which
substantially exceeds the 30% average over the 10-year period
in Figure 3. However, note that the total return for paired
switching is only double that for BAH of VFINX whereas, in
Figure 3, the total return for paired switching over the 10-year
period is three times that of SPY.

PAGE 94 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Figure 10. Maximum draw down on day N, the end day of

the 25-year period beginning on 1/2/1988 and ending on Finally, we should comment on the evaluation period, k. In
12/31/13 for entry date corresponding to day k, 1 ≤ k ≤ N this study we chose to evaluate the paired switching portfolio
= 6966 at the end of each month, in accordance with the rule used for
the Ivy 5. However, as mentioned earlier, Maewal and Scalaton8
only evaluated the possibility of switch every three months,
corresponding to the 3-month look back period. Moreover,
Cohn9 employed an evaluation period of one day, together
with a look back period of 84 days (it can be shown that, with
a one day evaluation period, the optimal performance has a
steep maximum about the 84-day look back period). Table 1
summarizes the performance for these three cases over the 10-
year period ending on 12/31/13.
Table 1. Total return, CAGR, maximum drawdown,
and total number of trades for three different paired
switching parameter sets for the 10-year period
beginning on 1/2/2004 and ending on 12/31/13 (N = 2517)
100 (N1V − 1) % (N1c ) % (N1δmax )% N
t
1

k = 1 mo, µ = 3mo 294.1 14.6 18.9 45

k = 3 mo, µ = 3mo 193.9 11.3 24.1 41
k = 1 day, µ = 84 days 327.8 15.5 17.1 189

Figure 11. Number of trades on day N, the end day of the

It can be seen that with k = 3 months, the number of trades
25-year period beginning on 1/2/1988 and ending on
12/31/13 for entry date corresponding to day k, 1 ≤ k ≤ N
is only slightly less than with k = 1 month, but the performance
= 6966
suffers significantly. Conversely if paired switching is evaluated
on a daily basis, with a look back period of 84 days, a roughly 1%
increase in CAGR is obtained, together with a small decrease
in maximum draw down. However, in this case, the number of
trades increases by a factor of four so that slippage becomes a
significant factor.

Conclusion
Paired switching refers to investing in one of a pair of
negatively correlated assets and periodic switching of the
position on the basis of the relative performance of the two.
The specific pair examined in this study is the S&P 500 index
and the index for U.S. treasury bonds with 20 or more years
duration. It was shown that using mutual fund surrogates,
VFINX for the former and VUSTX for the latter, and switching
between the two at the end of each month based on whichever
had the higher return over the past three months, a total return
of almost 2400% could have been achieved over a period of 25
years ending on 12/31/13, double that obtained by buying and
holding VFINX alone in this period. Moreover, this result would
have been obtained with significantly less risk; draw down was
only a third that with buy-and-hold. Unfortunately, short-term
CAGR results are shown in Figure 9. Over the entire set of trading restrictions on these mutual funds, still in place, would
entry dates in the 25-year period the minimum, mean, and have made it difficult, if not impossible, for the average investor
maximum values for paired switching are 10.2%, 15.2%, and to use this method. With the advent of exchange traded funds,
31.7%, respectively. These values are comparable to the values however, the game has changed. ETFs are readily available that
obtained with SPY and TLT in the 10-year period and shown track the S&P 500 index (SPY) and the long bond (TLT). The low
in Figure 4. The maximum draw down numbers in the 25-year electronic trading cost coupled with the huge daily volumes
period, shown in Figure 10, are also similar to those for the 10- provide the liquidity with which slippage can be minimized.
year period in Figure 5 and the number of trades shown in Figure Using the same paired switching rule just described it was
11, a little less than six per year on average, compares with the shown that the total return over the 10-year period ending
4 1/2 average number obtained from Figure 6. on 12/31/13 was almost 300%, three times higher than that
obtained by buying and holding SPY alone, and also three

IFTA.ORG PAGE 95
 IFTA JOURNAL 2021 EDITION

times higher than that obtained with the popular Ivy 5 method
described in “The Ivy 5 Strategy” above. And, as with the mutual
funds over the 25-year period, the draw down with paired
switching was only a third that with buy-and-hold of SPY. The
number of trades with paired switching over the 10-year period
was 45, compared with 70 for the Ivy 5, although the latter had
slightly lower maximum draw down (14.6% versus 18.8% for
paired switching).

Notes
1 Lewis A. Glenn is a Founding Partner and Chief Scientific Officer of
Creative Solutions Associates LLC, a private investment and wealth
management group.
2 Malkiel, B. G., “A Random Walk Down Wall Street” (2007), W. W.

Norton & Company

3 Faber, Mebane T., A Quantitative Approach to Tactical Asset Allocation

(February 1, 2013). The Journal of Wealth Management, Spring 2007.

Available at SSRN: http://ssrn.com/abstract=962461
4 Faber, M. T. and Richardson, E. W., The Ivy Portfolio—How to Invest

Like the Top Endowments and Avoid Bear Markets, John Wily &
Sons, 2009.
5 Glenn, L. A., Market Timing with Volatility, Active Trader, 10, No.

7, July 2009, pp. 28-31; see also Beat the Market—A Strategy for
Conservative Investors (December 12, 2008). Available at SSRN:
http://ssrn.com/abstract=1315533
6 Glenn, L. A., Market Timing using Exchange Traded Funds, The

Technical Analyst, July-Sept. 2010, pp.16-24. Available at SSRN:

http://ssrn.com/abstract=1591969; see also Playing Both Sides,
Active Trader, 11, No. 9, September 2010, pp. 44-49
7 Glenn, L. A., Adaptive Market Timing with ETFs (December 28, 2010).

Available at SSRN: http://ssrn.com/abstract=1732010

8 Maewal, Akhilesh and Scalaton, Joel B., Paired-Switching for Tactical

Portfolio Allocation (August 22, 2011). Available at SSRN: http://

ssrn.com/abstract=1917044
9 Cohn, Marc, Return Like a Stock, Risk Like a Bond: 15.5% CAGR

with 17% Drawdown, (February 23, 2014), Available at http://

seekingalpha.com/article/2041703-return-like-a-stock-risk-like-a-
bond-15_5- percent-cagr-with-17-percent-drawdown?ifp=0.
10 Henriksson, R. D. and Merton, R. C., On Market Timing and

Investment Performance. II. Statistical Procedures for Evaluating

Forecasting Skills, J. Business, 54, No. 4, October 1981, pp. 513-533.

PAGE 96 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

How the Average Investor Can Use Technical Analysis for Stock
Profits: An In-Depth Work on Stock Market Technical Analysis,
Mob Psychology, and Fundamentals by James Dines
Reviewed by Regina Meani, CFTe

Amazon describes it as James Dines’ famous treatise on same; they just look slightly different and are a little harder to
technical analysis for the stock market and advises that there manually construct.
are no reviews. On the Dines’ Letter website, there is a quote Dines manages to bring reality to his text by using excerpts
from a Barron’s Magazine reviewer, but I fear this was back in from the Dines Letter. In one interesting piece, he deals with the
the 1970s and no longer readily available. I reached onto my criticisms of the Dow Jones Industrial Average as to its value as
bookshelves and pulled out my treasured tome, thinking it a market indicator, which remains a relevant topic to this day.
was time for a 21st century review. First published in 1972 after For his argument, he uses quotes from the New York Times and
taking James Dines 11 years to write, he set about publishing from Arthur Merrill’s Behavior of Prices on Wall Street. The joint
it himself through the Dines Chart Corporation, and it was conclusion from the comparison to the broader indices was
reprinted in 1973 and 1974. Not only was he a pioneer in his that “The price action in the two averages has been amazingly
thinking on stock market analysis but also in publishing his similar. But this nonetheless does not prove the DJI is always
own work nearly 50 years ago. There is a massive 599 pages that representative of the market.”5
have been written to involve both the beginner and the more There are detailed descriptions of chart patterns, moving
advanced. The London Financial Times commented that “Mr averages, and both financial and technical indicators and, most
Dines has done for Technical Analysis what Graham & Dodd has importantly, how to use these to make profits. I continue to use
done for Fundamental Analysis.” many of his techniques, particularly those for moving averages,
Part One deals with Mass Market Psychology, and one is finding that along with his behavioural studies, they have stood
immediately confronted with the first topic: Sex and the secret the test of time and the changing market scene.
desire to lose. While some parts of this section may be questioned To continue a more detailed review, and do justice to Dines’
by modern society’s correctness, it remains an interesting work, would perhaps take most of the Journal pages, and
interpretation of people’s attitudes to the share market. We this would be without attempting to critique his section on
are also introduced to “Pigeon”, who features throughout and Fundamental Analysis. Many have referred to Dines’ book as a
I must admit has helped me out in many of my presentations. “classic” and to the man as a “legend”, and both no doubt have
Pigeon adds humour and allows us to connect with the text as we merit for maintaining positions of excellence for 50 years or more.
empathize with Pigeon’s ongoing predicaments. One of my more memorable moments was attending a talk by
2 Throughout, Dines tempers his work Dines when he visited Australia. I approached him, clutching my
with sage advice. By page 123, we have copy, and I believe that he was both pleased and surprised. I’m
reached Part Two—Technical Analysis. To not sure, but my guess was that he was surprised it had made
like everyone is to be indifferent to it to Australia due to its limited publication. I might also guess
everyone. Same with charts, you must that my signed copy would be worth more than the current price
choose. It is the act of choosing which is the of AUD$1,365 (US$994) and perhaps it could be said that it has
business of this book.3 proved one of my better investments.
In a clear and precise manner, Dines
explains the different types of charts in Notes
his section on “How to Chart”. He deals with Point and Figure, 1 The Dines Letter—Mr. Dines’ Bio: https://www.dinesletter.com/bio.
the arithmetic bar, and the semi-log bar and advises that we html
2 Dines, J, How the Average Investor Can Use Technical Analysis for
should use all three. He states that “Each measures different
technical factors: price change and volume in time, percentage Stock Profits, Dine Chart Corporation, USA, 1972, p. 1
3 ibid, p. 123
of price change and volume in time and price change alone. All 4 Ibid, p. 323
reveal different facets of a stock’s personality-in-motion. While the 5 Ibid, p. 345
methods of analysis are homologous. The conclusions reached are 6 Ibid, p. 138
often surprisingly different.”4 It is a shame that we do not have his
interpretation of candlesticks, but I believe that it wasn’t until
the 1990s that they became popular in the western world. Dines 6

uses Point and Figure charts for most of his prolific examples,
and he uses his own particular style. Most of us are familiar
with the “X” and “O” technique, but Dines just uses crosses
and refers to them as “Cross Charts”. The interpretation is the

IFTA.ORG PAGE 97
 IFTA JOURNAL 2021 EDITION

Author Profiles

Ron Albert Marcelino Acoba, CFTe, CMT, MFTA Lewis A. Glenn, Ph.D.
Ron Albert Marcelino Acoba, CFTe, CMT, MFTA, Lewis A. Glenn, Ph.D., is a founding partner and
is the chief investment strategist and co- chief scientific officer at Creative Solutions
founder of Trading Edge Consultancy, a third- Partners Associates LLC, a private investment
party research provider for banks and and wealth management group with offices in
brokerage firms in the Philippines. Ron has more the San Francisco Bay Area and Lausanne,
than 15 years of experience in the investment Switzerland. In a previous life Lewis led the
industry. He used to be an equity dealer and forex trader for Computational Physics Group in the Earth Sciences Division at
Credit Suisse Philippines, as well as an equity fund manager and the Lawrence Livermore National Laboratory in Livermore,
chief equity dealer for BPI Asset Management and Trust Group, California.
where he personally managed US$500 million in discretionary
funds. Earlier in his career, he worked as a forex analyst and Yukitoshi Higashino, MFTA
trader for BabyPips.com, LLC. Ron received his undergraduate Yukitoshi Higashino, MFTA, is chief technical
degree in economics from the Ateneo de Manila University. He strategist of the equity research team at DZH
also holds an MBA from the University of the Philippines. Financial Research, an investment information
firm located in Tokyo as Shanghai DZH Limited
Momen Atef El Shayal, CFTe, MFTA group. He leads the development of technical
Momen Atef El Shayal, CFTe, MFTA, is the head strategies for client securities firms dealing in
of the technical analysis department at SIGMA futures, ETFs, and CFDs. His main responsibilities range from
Securities Brokerage. He has almost 10 years of analysis of major indices of the Asian and world markets and
experience in the capital markets. He earned a can extend to even individual stocks. Prior to this, he was a
B.Sc. in commerce in 2008 and acquired his CFTe stock lending trader for foreign securities houses and a treasury
in 2013. Momen is a member of the Egyptian stock trader for Mizuho Trust Bank. He was also an equity
Society of Technical Analysts (ESTA). He started his career in trader and a market analyst at a securities firm. Yukitoshi is
Jazira Capital as a senior technical analyst in 2011 and was director of the Nippon Technical Analysts Association (NTAA).
promoted to chief technical analyst in 2013, serving both asset He is also a member of the Financial Planners Association of
management and brokerage divisions. In 2016, he joined the Japan (CFP). He made presentations at the IFTA conferences in
SIGMA Securities Brokerage team, where he is responsible for Vancouver, Lugano, Berlin, and Cairo.
technical analysis products for individual and institutional
clients covering the MENA and U.S. markets. Mohamed M. Khedr, CFTe, MFTA
Mohamed M. Khedr, CFTe, MFTA, has been
Mohamed Fawzy ElSayed Ali AbdAlla, CETA, CFTe, working as a senior technical analyst at Prime
MFTA Securities since 2014. He has over 12 years of
Mohamed Fawzy ElSayed Ali AbdAlla, CETA, experience in the Egyptian capital market.
CFTe, MFTA, is a urology consultant in the Before joining Prime Securities, he was a senior
National Institute of Urology and Nephrology in account manager at NAEEM Holding Co. Prior to
Egypt. He is also deeply interested and involved that, he worked as a senior technical analyst at Tycoon
in the analysis of financial markets. He has five Securities. Mohamed started his career at Mirage Brokerage Co.
years of experience as an independent in the customer service department. He then became a broker, a
investment strategist handling his family portfolio. research technical analyst, and finally the head of the technical
As a private investor, Mohamed grew to believe that analysis department. He also created a new department (call
achieving continuous portfolio growth is obtained by creating center) to answer clients’ technical questions. Mohamed has a
and improving a reliable money management trading system, bachelor’s degree in accounting from the Faculty of Commerce
which can only be reached through integration of two main at Cairo University. He has a brokerage license from ECMA
success tools: technical analysis as the only signal provider (Egyptian Capital Market Association) and is a CPM holder
and, more importantly, statistics and programming languages (Certified Portfolio Manager) from EIMA (Egyptian Investment
for better diversification, allocation, risk management, and Management Association).
performance analysis.
Kevin Luo
Profile not available.

PAGE 98 IFTA.ORG
 IFTA JOURNAL 2021 EDITION

Mohamed Ashraf Mahfouz, CETA, CFTe, MFTA Przemysław Smoliński, MFTA

Mohamed Ashraf Mahfouz, CETA, CFTe, MFTA, is A 20-year enthusiast of technical analysis and
chief technical analyst for Commercial trading, both professionally and as a hobby,
International Brokerage Company (CIBC), one of Przemysław Smoliński, MFTA, got interested in
Egypt’s top-ranked brokerage houses. CIBC is the stock market in the last year of primary
the securities brokerage division of CI-Capital, school and still has been fascinated by drawing
the premier investment bank in Egypt, with lines, designing, and coding trading strategies.
market-leading leasing, microfinance, investment banking, From 2000–2006, he worked at mBank Securities, and since
securities brokerage, asset management, and research 2006, he has been working as an analyst in the brokerage house
franchises. Prior to this he was the head of the technical of the largest bank in Poland, PKO BP—first in the Research
analysis desk for Dynamic Securities Brokerage Company, a Department and currently at the Technical Analysis and
CI-Capital member. Investment Advisory Department. Invariably, since 2012, he has
He served as a board member and vice president for the been considered one of the best technical analysts in the
Middle East and Africa region of IFTA from 2010 until 2016. newspaper’s “Parkiet” ranking, and his analytical products
During this period, he implemented MOU agreements between three times got to the short list of finalists in the competition
IFTA and the United Arab Emirates Securities and Commodities organized by the British magazine The Technical Analyst.
Authority (ESCA) and between IFTA and the Capital Market Nowadays, Przemysław is mainly focused on combining classical
Authority in Sultanate Oman (CMA), and he was responsible technical analysis and statistical tools with the possibilities
for translating the CFTe bank of questions into the Arabic offered by automated trading.
language. He also served as a board member in the Egyptian
Society of Technical Analysts (ESTA) from 2011 until 2015, and
he has been a key education committee member in ESTA since
2007. Mohamed has taught technical analysis at many financial
institutions and universities for more than 14 years. He
graduated from the Faculty of Economics and Political Sciences,
Cairo University.

Regina Meani, CFTe

Regina Meani, CFTe, covered world markets as a
technical analyst and associate director for
Deutsche Bank prior to freelancing. She is an
author in the area of technical analysis and is a
sought after presenter both internationally and
locally, lecturing for various financial bodies and
universities as well as the Australian Stock Exchange. Regina is
a founding member and former president of the Australian
Professional Technical Analysts (APTA) and a past journal
director for IFTA, carrying the CFTe designation and the
Australian AMT (Accredited Market Technician). She has regular
columns in the financial press and appears in other media
forums. Her freelance work includes market analysis, webinars,
and larger seminars; advising and training investors and traders
in market psychology; CFD; and share trading and technical
analysis. She is also a former director of the Australian
Technical Analysts Association (ATAA) and has belonged to the
Society of Technical Analysts, UK (STA) for over 30 years.

IFTA.ORG PAGE 99
 IFTA JOURNAL 2021 EDITION

IFTA Staff IFTA Board of Directors

Executive Director IFTA HEADQUARTERS
Charles W.L. Deale, FASAE, CAE International Federation of Technical Analysts
Senior Vice President, Meetings 1300 Piccard Drive, Suite LL 14
Grace L. Jan, CAE, CMP Rockville, MD 20850 USA
Phone: +1 (240) 404-6508
Senior Member Services Manager Fax: +1 (301) 990-9771
Linda Bernetich, CAE Email: [email protected] | Web: www.ifta.org
Production Manager
Maryia Alenchyk President
Wieland Arlt, CFTe (VTAD)
Managing Editor
Lynne Agoston Treasurer; Website Director
Karin Roller, CFTe (VTAD)
Director of Accounting Services
Dawn Rosenfeld Secretary; Membership Director
Alek Jankowski, BE, M.Eng.Sc., Grad.Dip.Mgt. (ATAA)
Vice President; Vice President, Asia-Pacific
Hiwon Yoon, Ph.D., CMTA (NTAA)
Vice President, Americas; Conference Director
Jeanette Young, CFP®, CMT, M.S. CFTe (AAPTA)
Vice President, Middle East and Africa; Immediate Past
President
Mohamed El Saiid, MFTA, CFTe (ESTA)
Vice President, Europe
Clive Lambert, MSTA, CFTe (STA)
Education Director
Saleh Nasser, CMT, CFTe (ESTA)
Examination Director
Gregor Bauer, Ph.D., CFTe (VTAD)
Marketing and Webinar Director
Thomas Hicks, MSTA (STA)
Development Director
Ron William, MSTA, CFTe (SAMT)
Director
Giovanni Trombetta, Electronic Engineer, CFTA (SIAT)
Director
Francesco Caruso, MFTA (SIAT)
Director
Nik Ihsan Bin Raja Abdullah, MSTA, CFTe (MATA)
Director
Yukitoshi Higashino, MFTA (NTAA)

PAGE 100 IFTA.ORG

See what
others don’t.
Bloomberg charts harness the full breadth and depth
of information on the Bloomberg Terminal®, creating
visualizations that uncover key insights — and help you
generate powerful ideas and strategies.

• Monitor market trends

• Validate your trading ideas
• Collaborate in real time
• Backtest & optimize strategies

bloomberg.com/professional

© 2020 Bloomberg L.P. 901478 1020

Certified Financial Technician (CFTe) Program
IFTA Certified Financial Technician (CFTe) consists of the CFTe I and CFTe II
examinations. Successful completion of both examinations culminates in the
Curriculum
award of the CFTe, an internationally recognised professional qualification in The CFTe II program is designed for self-study, however, IFTA will
technical analysis. also be happy to assist in finding qualified trainers. Local societies
may offer preparatory courses to assist potential candidates.
Syllabuses, Study Guides and registration are all available on the
Examinations IFTA website at http://www.ifta.org/certifications/registration/.
The CFTe I exam is multiple-choice, covering a wide range of technical
knowledge and understanding of the principals of technical analysis; it is To Register
offered in English, French, German, Italian, Spanish, Arabic, and Chinese; Please visit our website at http://www.ifta.org/certifications/
it’s available, year-round, at testing centers throughout the world, from IFTA’s registration/ for registration details.
computer-based testing provider, Pearson VUE.
The CFTe II exam incorporates a number of questions that require essay- Cost
based, analysis responses. The candidate needs to demonstrate a depth of IFTA Member Colleagues Non-Members
knowledge and experience in applying various methods of technical analysis. CFTe I $550 US CFTe I $850 US
The candidate is provided with current charts covering one specific market CFTe II $850* US CFTe II $1,150* US
(often an equity) to be analysed, as though for a Fund Manager.
*Additional Fees (CFTe II only):
The CFTe II is also offered in English, French, German, Italian, Spanish, Arabic,
$100 US applies for non-IFTA proctored exam locations
and Chinese, typically in April and October of each year.

Master of Financial Technical Analysis (MFTA) Program

IFTA’s Master of Financial Technical Analysis (MFTA) represents the highest Timelines & Schedules
professional achievement in the technical analysis community, worldwide.
Achieving this level of certification requires you to submit an original body of There are two MFTA sessions per year, with the
research in the discipline of international technical analysis, which should be following deadlines:
of practical application. Session 1
“Alternative Path” application deadline February 28
Examinations Application, outline and fees deadline May 2
In order to complete the MFTA and receive your Diploma, you must write
Paper submission deadline October 15
a research paper of no less than three thousand, and no more than five
thousand, words. Charts, Figures and Tables may be presented in addition. Session 2
Your paper must meet the following criteria: “Alternative Path” application deadline July 31
• It must be original Application, outline and fees deadline October 2
• It must develop a reasoned and logical argument and lead to a sound Paper submission deadline March 15 (of the
conclusion, supported by the tests, studies and analysis contained in the following year)
paper
• The subject matter should be of practical application To Register
• It should add to the body of knowledge in the discipline of international Please visit our website at http://www.ifta.org/certifications/
technical analysis master-of-financial-technical-analysis-mfta-program/
for further details and to register.
Cost
$950 US (IFTA Member Colleagues);
$1,200 US (Non-Members)