Excerpt from “Harmonic Trading Volume Three” (2016)
“Harmonic Pattern Size
Throughout the years, I have been frequently asked about the proper size for a pattern. It is quite common for smaller structures to form and yield reactive moves but the situations often do not possess the sheer size in terms of number of price bars or the entire range to be a profitable opportunity. As I previously stated, harmonic patterns are specific sets of data points that comprise a collective signal. However, the minimum size of the structure must be established to measure patterns of significance and ignore irrelevant formations.
Each individual price bar represents the recorded history of trading within a certain time interval. Although the completion of most patterns includes several price bars, there are definitive data points that define the opportunity. Further in this material, we will examine some precise strategies that capitalize on these individual price bars. For now, the general question of how many price bars any harmonic pattern should possess must be answered with respect to a basic scientific assessment of what constitutes a sufficient sample size. For statistical studies, most require a minimum sample size of 30 data points. Related to harmonic patterns, this means that each pattern should possess 30 price bars regardless of the timeframe. Founded in the principle of the Central Limit Theorem, the 30-sample size is designed to ensure a normal distribution of probable outcomes. That is, we can expect a normal reaction at the completion of harmonic pattern opportunity more so in the structures then those that possess less than the ideal sample size amount.
In The Harmonic Trader, I established basic guidelines about pattern sizes. For example, I put forth the notion that the larger the pattern the more significant possible reversal. In another instance, a market opportunity that has multiple calculations converging in a specific area will likely favor the largest measured price leg is the most significant over smaller sized formations. Those are valid. However, I feel it is important to establish a standardize means of determining the proper sample size for harmonic patterns. If we postulate that 30 price bars will satisfy the requirements of normal expectations, it serves us to this those opportunities with at least this amount if not more. The problem arises when multiple time intervals have the same pattern. For example, a pattern that forms on a 60-minute chart will also be visible on a 15-minute chart. Of course, the 15-minute chart will have four times the number of hourly bars. If the hourly chart has 20 bars, the 15-minute chart would be favored having 60 bars and being closer to the 30-statistical minimum. However, when an hourly chart has 30 bars or something close to the minimum, sometimes the 15-minute chart more accurately reflects the possibilities of the situation. Simply stated, it is important to utilize that timeframe which is closest to 30 bars as the primary time interval to analyze parameters of the harmonic pattern.
Central Limit Theorem
I personally feel that the future of all market approaches will be optimized statistical analysis that emerges as the primary source of decision variables. As technological improvements put more power into the hands of the individual trader, a more scientific approach to assess market strategies. I have presented some research regarding harmonic phenomenon financial markets previously but I have withheld a great deal of statistical discoveries because these findings represent the ultimate realization of decades of work. Although one does not need to be a PhD statistician to incorporate these concepts into your decision framework, it is essential to understand principles of statistical probability as it relates to trading behavior. Ultimately, we are all executing definable market conditions that will someday be quantified and categorized into mathematically proven processes.
Since we have established harmonic patterns as a collection 30 or more individual price bars that possesses prescribed ratio proportions, we can focus on distinct price zones where normal expectations of possible outcomes can be realized. Simply stated, this means that patterns and harmonic measures consider will rarely get “blown out of the water” reducing vulnerable conditions that may arise from more random structures. Therefore, the principles of the Central Limit Theorem provide the basis of establishing some basic expectations of price behavior.
According to Wikipedia, “…the Central Limit Theorem (CLT) states that, given certain conditions, the arithmetic mean of a sufficiently large number of iterates of independent random variables, each with a well-defined (finite) expected value and finite variance, will be approximately normally distributed, regardless of the underlying distribution. To illustrate what this means, suppose that a sample is obtained containing a large number of observations, each observation being randomly generated in a way that does not depend on the values of the other observations, and that the arithmetic average of the observed values is computed. If this procedure is performed many times, the Central Limit Theorem says that the computed values of the average will be distributed according to the normal distribution (commonly known as a ‘bell curve’).“
Theory of Normal Distribution Returns
All patterns should possess at least 30 bars on the timeframe that they are identified. In fact, I find myself utilizing the 30 number to approximate a sufficient number of bars can encompass any trading situation. As I stated before, the pattern that is closest to 30 bars on any timeframes should be considered the most suitable for the primary pattern. Since all patterns serve as the basic measuring basis for entry points and stop loss considerations, we can assess the completion of patterns within this 30-bar limit as a minimum time constraint to allow the reversal to take hold. There are other examples where this approximation does work but the principle is founded in expectation of normal price behavior no matter what.
Albeit quite simple, this 30-bar limit has helped to confirm smaller patterns on shorter timeframe for that effectively optimize entries of larger structures. It is only in the past several years that I have been able to effectively explain the need for a minimum pattern size. However, as I have experimented with its application over the years, I know that this basic rule of 30 price bars has been quite effective as a means in differentiating those structures that are smaller and insignificant. If you are a long-term trader, I believe it is important to be aware of the last 30 days at 30 months no matter what the market. If you are a short-term trader, I would suggest always be aware of the prior day’s action – that is, the last 30 hours. Finally, the simple technique is yet another important principle to effectively analyze and monitor pattern progress.”