# Harmonic Pattern Book

The concept of trading with patterns has been around the financial market over 80 years even before R. N. Elliott pointed out the Elliott Wave patterns from US Stock market in 1940s. Since then, various patterns have been utilized by Forex and Stock market traders including Japanese candlestick patterns, Harmonic patterns, Breakout patterns, Elliott Wave patterns, etc. In this book, we mainly focus on the Harmonic patterns due to its wide usage in the financial market. The main idea behind the Harmonic pattern is that price movement is based on the structured relationship defined by Fibonacci ratios. The typical Harmonic patterns are constructed from 4 to 5 price points in your chart. The main application of Harmonic pattern is to identify the potential key turning point in the financial market. Harmonic patterns are typically accomplished by reversal traders. Due to the non-lagging advantage over the technical indicators, Harmonic Patterns are used by many professional traders in forex, future, and stock markets.

The history of the harmonic pattern goes back to the Gartley’s book “Profits in the Stock Market” in 1935. At that time, Gartley described the trend reversal pattern on page 222 of his book. The pattern become popular in the late 1990s (Pesavento and Shapiro, 1997). Carney (1998) identified the structure of the several popular harmonic patterns in his work. Since then, many traders developed the common interest in looking for the repeating patterns in the financial markets. The harmonic patterns were refined many times in several decades. Harmonic trader emphasizes that the patterns are not only repeating in history but they also follow natural orders. Although few different references exist about the meaning of the natural orders, the natural orders mostly means the periods of the neighbouring waves in the Fibonacci relationship (Pesavento and Shapiro, 1997). Fibonacci ratio derived from Fibonacci numbers are the core relationship used in harmonic pattern identification. To have a feel about the Fibonacci ratios, here is the 21 Fibonacci numbers derived from the relationship: Fn = Fn-1 + Fn-2. Many traders uses these Fibonacci number sequences to derive some important Fibonacci ratio.
0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89,144, 233, 377, 610, 987, 1597, 2584, 4181, 6765, …………………
As the Fibonacci number become large, the constant relationship is established between neighbouring numbers. For example, every time, when we divide the former number by latter: Fn-1/Fn, we will get nearly 0.618 ratio. Likewise, when we divide the latter number by former: Fn/Fn-1, we will get nearly 1.618. These two Fibonacci ratio 0.618 and 1.618 are the very important basis for the harmonic pattern. These two ratios are even used by many Elliott Wave practitioner too. In addition, harmonic pattern traders like to use 0.786 and 1.272 ratio. The ratio 0.786 and 1.272 are derived from the square root of 0.618 and 1.618 respectively. Likewise, one can derive the secondary ratio 0.382 and 2.618 by squaring 0.618 and 1.618. Many more secondary ratios can be derived using mathematical operations like square rooting, squaring, inversing, adding and subtracting the primary ratio 0.618 and 1.618. In addition, primary Fibonacci numbers like 2, 3, 5 and 13, which have no positive divisor other than one and itself, can be used to further generate the secondary ratios. In Table 1-1, we list some commonly used Fibonacci ratios and their calculation generally used by trader.

Type, Ratio, Calculation
Primary, 0.618, Fn-1/Fn of Fibonacci numbers
Primary, 1.618, Fn/Fn-1 of Fibonacci numbers
Primary, 0.786, 0.786=√0.618
Primary, 1.272, 1.272=√1.618
Secondary, 0.382, 0.382=0.618*0.618
Secondary, 2.618, 2.618=1.618*1.618
Secondary, 4.236, 4.236=1.618*1.618*1.618
Secondary, 6.854, 6.854=1.618*1.618*1.618*1.618
Secondary, 11.089, 11.089=1.618*1.618*1.618*1.618*1.618
Secondary, 0.500, 0.500=1.000/2.000
Secondary, 1.000, Unity
Secondary, 2.000, Fibonacci Prime Number
Secondary, 3.000, Fibonacci Prime Number
Secondary, 5.000, Fibonacci Prime Number
Secondary, 13.000, Fibonacci Prime Number
Secondary, 1.414, 1.414=√2.000
Secondary, 1.732, 1.732=√3.000
Secondary, 2.236, 2.236=√5.000
Secondary, 3.610, 3.610=√13.000
Secondary, 3.142, 3.142 = Pi = circumference /diameter of the circle

Table 1-1: Fibonacci ratios and corresponding calculations to derive each ratio.

Pattern Completion Interval is the emerging concept first introduced from this book. The concept was born after the extensive computerized research in tradable patterns in the financial market conducted by myself. Therefore, not many traders are aware of its existence yet. As you read this book, you will find out that it is extremely useful concept for your harmonic pattern trading. At the same time, the concept is not a rocket science. The concept is simple enough for any average trader for their practical trading. To understand the concept of the Pattern Completion Interval, we shall understand the term approximation first. Of course, everyone know the literal meaning of approximation. However, technically speaking, approximation make quite big influence every day in our life, but many people will not notice its impact unless you are the math geek crunching numbers all day in your job.

Whether you agree or not, approximation arises naturally every day in our life. There can be plenty of examples but we shall start with most intuitive one. Let us use the sprint record of Usain Bolt to learn about approximation. The record-breaking sprint of Usain Bolt was the popular coverage in many Newspapers during the 2016 Olympics since he was winning his third gold medal in 100 meter sprint. Here is simple but interesting three numbers about Usain Bolt, the fastest man in the world.

• Height: 1.96 meters (6 foot 5 inches)
• Distance: 100 meters
• Time: 9.58 seconds

These three numbers can be true but maybe not. I am not suspecting about legitimacy of Usain Bolt’s record like the drug test results in the Olympics. I am pointing out that the measuring instrument, whoever measured, can only approximately measure these numbers up to certain degree. It is not because the measuring person did his job poorly but just because the instrument have own limitation to measure these numbers. For example, the sprint time might be 9.5823 seconds instead of 9.58 seconds. Maybe expressing it into 9582.3 milliseconds, we can be slightly more precise. However, still we are not dead accurate. To be dead accurate, we need infinite number of decimals to describe these numbers. This is impossible. Most of time, we will always approximate regardless of what measurement unit we are using. Likewise, the height of Usain Bolt is only the approximation too. Precisely speaking it is impossible to tell if he is 1.963 meter tall or 1.962 meter tall. Besides the height and time approximation, you can probably find many other approximation examples in our daily life like weight, speed, calories, etc.

Here is another example. From your High School, you will probably remember pi, the ratio of a circle’s circumference to its diameter up to 2 decimal places as 3.14. Once again, this is only approximation. Some scientists remember it up to five decimal places as 3.14159 if they work frequently with pi. In fact, even if we use 50 decimal places to describe it as:
pi = 3.14159265358979323846264338327950288419716939937510,
we are only approximating it. By now, you should realize that countless approximation influence in and out of your life. One negative example might be that my classmate in my old university in the United Kingdom, failed to achieve the First Class honor since his overall score was only 69.4. In British degree system, First Class honor is granted to the students achieving the overall score over 70. First class honor is the highest grade they can achieve under the British degree system. At the same time, the other mate scored 69.6 earned First Class honor. Apparently, the academic satisfaction for these two friends were very different. Even after graduate, when they find jobs or when they get married, when they do business, these Second Class and First Class label will definitely stick with them. Now you can probably imagine that our world is not as pretty and square as you think. Well the same thing goes to scanning of Harmonic Patterns from your chart too.

Pattern Completion interval build its concept over the approximation but nothing else. It is in fact based on the assumption that the measured ratio in the harmonic patterns are only an approximation. It is because we assume that harmonic pattern should have the exact Fibonacci ratio in theory. However, when the Gartley pattern is detected by the pattern scanner, most of time the pattern will possess the approximated ratios, which closely match to the ideal Gartley pattern but not dead accurate. Well, one day you can be very lucky to find the perfect Gartly pattern with perfect ratio in your chart. This is very rare event. Even in this very rare event, the chance that your pattern will be truly perfect is very thin because the pattern scanner might round up the ratio AB/XA for 0.618 instead of 0.6181 or CD/BC for 1.272 instead of 1.2723. Approximation error is always there in our pattern detection task. We will never be able to get rid of them since we have only limited memories inside both human brains and computers.

Since this approximation exists every time in detecting harmonic pattern, we know that, we are less accurate every time when our pattern scanner measure the ratio 0.618 or 0.382 or other Fibonacci ratio from our chart. Well, this is very common facts in the scientific world. On the other hands, as this is so common, the scientist already gave a lot of thought in overcoming this approximation error rather than using infinite number of decimals to describe the measurement.

So how can we be overcome this approximation error? Typically, in practical application like engineering and statistics, people use tolerance as one possible way of describing the measurement. In technical term, tolerance is the total amount by which a specific dimension is permitted to vary. The tolerance is the difference between the maximum and minimum limits. Going back to our Usain Bolt’s sprint record. Instead of writing 9.58 seconds, we can write 9.58 seconds ± 0.005. This means that Usain Bolt’s record will not be greater than 9.585 seconds and it will not be smaller than 9.575 seconds. His record will fall somewhere in between 9.585 and 9.575. By assigning maximum and minimum tolerance limit, we can be more precise in recording his records. We can also avoid using the infinite decimal places to describe his record. Using infinite decimal places is impractical. Likewise, we can describe his height as 1.96 ± 0.005. This means that his height will fall in between 1.965 and 1.955 meters.

How this tolerance can be related to the Pattern Completion Interval in our Harmonic Pattern trading? Well, Pattern Completion Interval is in fact no more than just the tolerance limit described above. It is indeed the upper and lower limit permitted to vary in detecting Harmonic Pattern. Since detecting Harmonic Pattern is quit visual task, it might be a good idea to show the pattern completion interval using a box like in Figure 2-2. In the AB=CD Bearish reversal Pattern in Figure 2-2, the upper limit is the maximum price level permitted for this pattern to be qualified as AB=CD Harmonic Pattern. If EURUSD goes beyond this Upper Limit, then the Pattern can not be qualified as the AB=CD pattern since the pattern is breaching the tolerance limit for the given Fibonacci ratio.

In general, the tolerance limit in many practical applications are specified in symmetric manner like 1.96 meters ± 0.005. Technically, we can assign symmetric Upper and Lower Limit for Pattern Completion Interval too. However, either one limit between Lower Limit and Upper Limit is relevant for our trading depending our trading direction. For example, for Bearish Reversal Pattern, we only need to concern about Upper Limit since we want to know when the Harmonic Pattern will fail to form from the price moving too high. Likewise, for Bullish Reversal Pattern, we only need to concern about Lower Limit.

Another peculiar thing to discuss is that unlike our height and time example from Usain Bolt’s sprint, our pattern detection task involves several ratios and not just single ratio. For the Bearish AB=CD pattern, we will measure the ratio BC/AB and CD/BC together to check if these ratio are held near 0.618 and 1.271 respectively. Alternatively we can also check if the ratio is held near 0.786 and 1.618 respectively. Since we are checking two ratios, in fact we are concerning about 4 points for our pattern detection in our case of AB=CD pattern detection. Although the most important tolerance limit will be based on the final point D, one can still have the tolerance limits at the point A, B and C too. However, the tolerance limits at the point A, B and C are much less useful for our trading. In our operational definition, the Pattern Completion Interval is the tolerance limit at the final point D.

Pattern Completion interval and tolerance concept should clear up the most commonly asked questions from junior traders like:
Why detected Harmonic Patterns have the ratios different from the ideal one?
I am seeing the wrong patterns in my chart. Can I only see the ideal patterns in my chart?
My pattern is dead accurate. If I trade with this pattern, can I have success rate of 96%?

Although we use the analogy of tolerance to illustrate the concept of the Pattern Completion Interval. In the course of our research, the concept of pattern completion interval was actually drawn from the concept of the prediction interval and confidence interval from statistics. For simple example, consider the simple regression with prediction interval and confidence interval. The fitted line on the constructed regression equation is the point prediction for the given x values. If we are using this regression equation to predict future y values according to the given x values, you can see that the point prediction from the given fitted line will not accurately predict the future outcome. In fact, most of time, prediction from regression equation will offer us some reference points to consider but they will be quite far out from actual future outcome.

Instead of relying on this point prediction, statistician use prediction interval to illustrate worst and best case of your prediction in probabilistic sense. For example, 95% prediction interval means that future outcome will fall within this interval 95% of time. With the prediction interval, it is much easier to assess your risk of making wrong prediction than just using the point prediction alone. Likewise, the confidence interval is the similar concept to prediction interval. Instead of concerning the point prediction, Confidence interval concern our choice of parameters in our mathematical model. Since we can only estimate our parameters based on our sampled data, confidence interval help us to assess our risk associated with our choice of parameters. As before, 95% confidence interval means that the true population parameters will fall within this interval 95% of time. For both prediction interval and confidence interval, 90%, 95% and 99% are the common intervals to use. However, some industries use some other intervals like 80%.

We found that our Pattern Completion Interval shares many common ideas with prediction Interval and confidence interval. However, it is difficult to tell which one is closer to Pattern Completion Interval. The Pattern Completion Interval helps us to assess the risk associated with the Harmonic Pattern formation. If we view the harmonic pattern detection task as the prediction of turning point, then the Pattern Completion Interval offer the similar functionality to prediction interval to users. However, the Pattern Completion Interval does not offer us any probabilistic information like prediction and confidence interval do. Since we assume the Fibonacci ratio as an absolute measure for the Harmonic Pattern formation, the typical concern rising over the choice of parameters, which is necessary for many other practical applications, fade away. Since this book focus on the practical side, it make more sense to concern about the functionality of Pattern Completion Interval for traders.

## List of Harmonic Pattern Trading Books

Here is the list of Harmonic Pattern Book available for Forex and Stock Market Trader. These two book will provide the great deal of information across Fibonacci Analysis, Harmonic Pattern, Elliott Wave, and X3 Chart Pattern in Forex and Stock Market Trading. If you are day trader, then you must read these two books. These books are available from all the major book distributor including Scribd, Google Play Book, Apple Book, Amazon.com and so on.

1) Guide to Precision Harmonic Pattern Trading (Mastering Turning Point Strategy for Financial Trading)

2) Profitable Chart Patterns in Forex and Stock Market (Fibonacci Analysis, Harmonic Pattern, Elliott Wave, and X3 Chart Pattern)

https://www.amazon.com/Kindle-Store-Young-Ho-Seo/s?rh=n%3A133140011%2Cp_27%3AYoung+Ho+Seo

## List of Harmonic Pattern Indicator

Below is the website link providing powerful Harmonic Pattern and Elliott Wave Pattern Trading tools for Forex and Stock Market.

a) Harmonic Pattern Plus

Harmonic Pattern Plus is the harmonic pattern scanner. It can detect harmonic patterns with predefined trading entries and stops. Pattern Completion Interval, Potential Reversal Zone and Potential Continuation Zone are supported. In addition, other dozens of powerful features are added like Japanese candlestick detection, automatic channel, etc.

https://www.mql5.com/en/market/product/4488

https://www.mql5.com/en/market/product/4475

b) Harmonic Pattern Scenario planner

With additional features of predicting future harmonic patterns, this is the tactical harmonic pattern indicator with advanced simulation capability on top of  many powerful features of harmonic pattern plus. Try it then you will know its power.

https://www.mql5.com/en/market/product/6240

https://www.mql5.com/en/market/product/6101

c) X3 Chart Pattern Scanner

X3 Chart Pattern Scanner is non repainting and non lagging pattern scanner for Harmonic Pattern, Elliott Wave pattern and X3 patterns. It is the most advanced but user friendly harmonic pattern indicator. As a bonus, this tool can detect around 52 bearish and bullish Japanese candlestick patterns + advanced Channel features. This is the most advanced Harmonic and Elliott wave scanner in the market. You must try.

https://www.mql5.com/en/market/product/41993

https://www.mql5.com/en/market/product/41992