| Current Members Log-In | |
View Your Shopping Cart |
CRB Bookstore |
Markets Overview |
CRB Affiliates |
|
![]() |
| |
|
|
Publications |
|
|
|
|
By Bernard Mitchell Leonardo Fibonacci, a mathematician, created a numerical sequence of numbers while studying the Great Pyramid of Gezah. The fascinating concept that Fibonacci discovered is that as the sequence of numbers proceeds, each number in the sequence is 1.618 times the preceding number and .618 of the next number. (1, 2, 3, 5, 8, 13, 21, 34, 55, 89, 144, 233, 377, 610, 987 etc.) The sequence of numbers is believed to occur throughout nature. Many technicians use the Fibonacci numbers when trying to determine support and resistance, and commonly use .382, .50, and .618 retracements. It is commonly believed that, a .382 retracement from a trend move will tend to imply a continuation of the trend. A .618 retracement implies that a trend change may be in the making. Many such rules have been adopted by technicians. My approach is different in that we do not use Fibonacci retracement numbers to determine the trend. My long-term trend is determined by a simple 89-period moving average. It is effective from a one-minute chart to a monthly chart. This moving average is the literal battleground between the bulls and bears. My use of Fibonacci Clusters will determine support and resistance in the direction of the main trend (which is always determined by prices either above or below the long-term moving average)—what I call our "Main 89 line." Most charting packages will enable you to create a simple 89-period moving average. I encourage you to look at the market this way. What are Fibonacci Clusters and how are they different in calculating support and resistance than the standard ratios with which most traders are familiar? Simply put, Fibonacci Clusters are more accurate because of the use of more data points and the way the Fibonacci ratios are calculated. I include as many as seven swing highs and swing lows in addition to the significant high points and significant low points in the calculations. Let’s review the approach I use to calculate support and resistance levels. The first thing we need to identify is a trend move, i.e. a significant high point to a significant low point or vice versa. These points are the extreme boundaries of the trend. Within these extreme points are contra-trend swing highs within an overall downtrend and swing lows within an overall uptrend. In other words, I look to "capture" the high/ low points of a contra-trend move within the longer trend. These contra-trend high/low points are swing points and are critical to my calculation of Fibonacci Clusters. The normal retracement levels of .382, .5, and .618 of the maintrend, are simply not accurate enough when the pinpointing of support and resistance becomes crucial to pulling money out of the markets. First, it’s important to understand the concept of support and resistance. In this writer’s opinion, unless you can determine precise support and resistance, most traders will achieve mediocre results at best. Simply put, support means there is sufficient buying pressure (volume) at a given price level to halt a downtrend and resistance means there is sufficient selling pressure (volume) at a given price level to halt an uptrend. Support places a floor under the market and resistance places a ceiling above the market. The higher the volume, relative to the range of the price bar, the more significance is given to the price level as support and resistance. I measure this in terms of ticks in a price bar. A greater number of transactions (ticks) within a given time frame—and this can be anything from a one-minute price bar to a monthly bar—when divided into the range of the bar and compared with the previous two price bars, with the mathematical result being lower, will produce a robust squat with high predictive power for market reversals. The squat bar is my trigger to enter a trade. The only thing I now need to know is where support and resistance will likely take place. And, that is where my calculation of Fibonacci Clusters comes in. The resulting output is four levels of support and four levels of resistance. They are:
Each of first three levels of resistance and first three levels of support are tradeable levels, as long as a squat is generated on the entry bar. The maximum resistance and support levels are normally not used for entries—unless they are a "confluence" number from a larger or smaller trend. More on this later. Prior to the development of software, it was too time consuming to create these levels of support and resistance, except on daily and weekly data. We can now produce reliable, tradeable numbers on intra-day data even when highs and lows are constantly changing. The reader can test the efficacy of Fibonacci Clusters, by requesting a demo version of my software. To Find Resistance Observing the chart, you can easily identify the highest high (SHP) and the lowest low (SLP) in the trend and a series of swing highs (SH) that fall between the two extreme points. You need to identify these prices.
The selection of swing highs is straightforward. There needs to be a minimum of three bar highs. When this procedure is completed, activate the calculation by clicking the "RESISTANCE" button. To Find Support Observing the chart, you can easily identify the highest high (SHP) and the lowest low (SLP) in the trend and a series of swing lows (SL) that fall between the two extreme points. You need to identify these prices.
The selection of swing lows is straightforward. There needs to be a minimum of three bar lows. Let’s now look at some examples of Fibonacci Clusters and apply the calculations to different time frames, i.e. weekly, daily and intra-day. Weekly—ATML CORP (ATML)
Note that the market then proceeded to stage a choppy rally over 10 weeks, peaking at 18.438; just shy of our stopping point, 18.62. Over the next 10 weeks a new low was made at 7.625. Using the new low as my new SLP, and using the 18.438 rally high as my new SH3, a new series of resistance numbers have now been projected. Since this is in the future, we need to wait and see how these numbers eventually work out. Observing the output from the two trends, each with a different SLP, i.e. 9.375 and 7.625 a series of confluence resistant points have shown up. They are:
At these three levels, shorts, or exits to previously held long positions can be taken, provided a squat bar is generated on the daily charts. Daily—Texas Instruments (TXN)
Going one step further in attempting to find confluence, I used the SH1 as my SHP. The result was:
You will note that the lower resistance (32.94) on the longest trend comes close to the stopping point (32.49) on the shorter trend and stopping point on the longer comes fairly close (not perfect) to upper resistance on the shorter trend. Observing the chart, the high print of 34.01 falls between these two numbers, before the market started a significant sell off. WorldCom (WCOM) To find confluence, I used the SLP of 16.125 with the same SHP. This produced important support between 17.37 to 17.28. As you can see, the market spent 26 days trying to hold this support level, before finally succumbing in a waterfall decline to 13. Intra-day—September S&P Futures 135-minute Chart
From the SLP low, the market rallied to 1244.90 eight bars later, just shy of lower resistance. A series of squat bars showed up as the market spent nine more bars before finally tanking. Confluence numbers were also created showing important resistance around 1244. Now, (July 6) with new lows being made a new series of resistance levels can be calculated. E- Mini S&P—610 ticks I have created resistance levels from the huge downtrend that started with the SHP at 1247.25 on July 3 at 6:00 AM E.T., culminating in a SLP at 1189 on July 9 at 10:04 AM E.T. Using seven swing highs as my inputs along with SHP and SLP the resistance levels are as follows:
In addition, I have attempted to find confluence numbers by using two different SHPs, thus calculating from smaller trends. Those resistance levels come in at:
The very first thing I notice are numbers that are confluencing around 1214.04 (lower resistance on the largest trend). They are 1214.44 (stopping point next largest trend), and 1214.35 (upper resistance smallest trend). I will average these three numbers and arrive at 1214.27. The series of confluence numbers on the big trend come in around 1220.75. They are 1219.81 (stopping point largest trend). Those numbers average out to 1220.82. The third series of confluence numbers on the big trend come in around 1228.25. They are 1226.45 (upper resistance largest trend) and 1229.87 (maximum resistance smaller trend). They average 1228.16. Shorts can be taken at all three levels as long the entry bars are accompanied by squats. The thing to remember: this resistance will be constantly updated in real time as new lows are made. After more than 20 years of working with the natural numbers that make up the Fibonacci sequence, Fibonacci clusters with the power of confluence come closest to pinpointing support and resistance. Bernard Mitchell is the president of PBSP LLC. He started trading full-time in 1981 when he developed the "POKE" system for the commodities markets. He modified the same system for day trading in 1994 focusing on the S&P 500 futures and then for individual stocks in 1996. He holds a Series 3 License and manages a branch office of the Peregrine Financial Group, a commoditiy broker based in Chicago. He is a systems developer with designation as an Omega Research Solutions Provider under PBSP LLC, which provides trading advice over the Internet and educational programs. He was a beta test site for Omega Research’s first product, System Writer in 1986, and was a beta site for the new TradeStation 2000i. He is on the graduate level faculty of E-CADEMY of trading. Based in Chicago, E-CADEMY presents classes and workshops nationwide. Mitchell has successfully trained more than 300 traders on the "POKE" system and conducts online trading seminars which can be found on www.feargreed.com.
CRB TRADER is published bi-monthly by Commodity Research Bureau, 330 South Wells Street, Suite 612, Chicago, IL 60606-7110. Copyright © 1934 - 2002 CRB. All rights reserved. Reproduction in any manner, without consent is prohibited. CRB believes the information contained in articles appearing in CRB TRADER is reliable and every effort is made to assure accuracy. Publisher disclaims responsibility for facts and opinions contained herein. | |||||||||||||||||||
|
|
Copyright © 1934 - 2008 by Commodity Research Bureau - CRB. All Rights Reserved.
User agreement applies. Privacy policy. 330 South Wells Street Suite 612 Chicago, Illinois 60606-7110 USA Phone: 800.621.5271 or 312.554.8456 Fax: 312.939.4135 Email: info@crbtrader.com |
| Press Ctrl+D to bookmark this page - Set http://www.crbtrader.com as your Home Page |