The Elliott Wave Fractal Model

Elliott Wave Fractal Model (a.k.a. “The Wave Principle”) for forecasting Price Direction and Trend Change Levels

Background: “The Wave Principle” is Ralph Nelson Elliott’s 1930’s discovery that social, or crowd, behavior trends and reverses in recognizable patterns. Using stock market data as his main research tool, Elliott discovered that the ever-changing path of stock market prices reveals a structural design that in turn reflects a basic harmony found in nature (Fibonacci mathematics). From this discovery, he developed a rational system of market analysis. Elliott isolated thirteen patterns of movement, or “waves,” that recur in market price data and are repetitive in form, but are not necessarily repetitive in time or amplitude. He named, defined and illustrated the patterns. He then described how these structures link together to form larger versions of those same patterns, how they in turn link to form identical patterns of the next larger size, and so on. In a nutshell, then, the Wave Principle is a catalog of price patterns and an explanation of where these forms are likely to occur in the overall path of market development, i.e. a fractal model of the markets. Elliott’s descriptions constitute a set of empirically derived rules and guidelines for interpreting market action.
Basic Model description: In markets, progress ultimately takes the form of five waves of a specific structure. Three of these waves, which are labeled 1, 3 and 5, actually effect the directional movement. They are separated by two countertrend interruptions, which are labeled 2 and 4, as shown below (Figure 1). The two interruptions are apparently a requisite for overall directional movement to occur.

Figure 1

At any time, the market may be identified as being somewhere in the basic five wave pattern at the largest degree of trend. Because the five wave pattern is the overriding form of market progress, all other patterns are subsumed by it. There are two modes of wave development: motive and corrective. Motive waves have a five wave structure, while corrective waves have a three wave structure or a variation thereof. Motive mode is employed by both the five wave pattern shown above and its same-directional components, i.e., waves 1, 3 and 5. Their structures are called “motive” because they powerfully impel the market. Corrective mode is employed by all countertrend interruptions, which include waves 2 and 4 in the above figure. Their structures are called “corrective” because they can accomplish only a partial retracement, or “correction,” of the progress achieved by any preceding motive wave. Thus, the two modes are fundamentally different, both in their roles and in their construction.  One of the major contributions of the Wave Principle is an arguably objective determination of whether a market is in a motive or corrective wave at the observed degree of trend.
Any market unfolds according to a basic rhythm or pattern of five waves up and three waves down to form a complete cycle of eight waves. The pattern of five waves up followed by three waves down, as well as the idealized fractal nature of the patterns is shown in the figure below.  The market’s compound construction is such that two waves of a particular degree subdivide into eight waves of the next lower degree, and those eight waves subdivide in exactly the same manner into thirty-four waves of the next lower degree. The Wave Principle, then, reflects the fact that waves of any degree in any series always subdivide and re-subdivide into waves of lesser degree and simultaneously are components of waves of higher degree. Thus, we can use Figure 2 below to illustrate two waves, eight waves or thirty-four waves, depending upon the degree to which we are referring.

Figure 2

Rules of Wave formation:  There are 3 rules and a strong guideline (sometimes called the 4th rule) that the 8 wave Elliott structure for a bull and bear market follow and numerous guidelines based on a comprehensive study of all historical price data.
      1. Wave 2 does not retrace below the start of wave 1
      2. Wave 4 does not enter into the price territory of wave 2 for impulsive motive waves.  In the case of diagonal motive waves, there can be an overlap.  However, the structures of an impulsive and diagonal motive waves are different enough to eliminate any ambiguity.
      3. Wave 3 is never the shortest wave of 1, 3, and 5.
      4. Waves 2 and 4 alternate in corrective form, i.e. if wave 2 is a zig-zag, then wave 4 is a flat correction and vice versa.
Wave Character:  Each wave has a character to it that is corroborated by coincident economic data (if the timeframe being observed is long enough) and/or by other technical indicators like price momentum, breadth, volume, and sentiment.  For example 3rd waves tend to be the strongest in these terms with 5th waves almost always being weaker than 3rd waves even if they are longer in absolute price movement.  We use this characteristic of waves in conjunction with proprietary indicators and methods to identify the likely wave counts for an asset.
Fractal similarity:  An important corollary of Wave Character is that waves of a similar numbering can to be similar in form even at different degrees of trend.  i.e. a wave (5) of 3 in the above figure would be similar to a wave (5) of 5, or even similar to the entire wave 5 itself.  This stems from the inherent fractal nature of asset prices.  But the beauty is that if underlying events rhyme then the patterns tend to be similar as well.  The first example shown in Figure 3 compares the Dow period from 1921 to 1929 to the Dow period of 1974 to 2000.  Both were 5th waves as shown within the larger historical context in Figure 2 and both were driven by historically unprecedented credit bubbles.

Figure 3

The second example is shown in Figure 4 which compares a Nasdaq rally from 2003 to 2007 (which is clearly a corrective rally in a primary bear market) to the recent rally in the Nasdaq from March 2009 to April 2010 (which we have also classified as a corrective rally within the context of a now secular bear market) with data continuing on till August 2010.  Context is provided in Figure 3 where the Nasdaq’s rise from 2003 to 2007 is a wave B of an A-B-C decline that started in 2000 and the Nasdaq’s rise from March 2009 to April 2010 is shown as a wave (2) of C of the same secular bear market.  Both declines are considered a triple-zig-zag, which is one of the 11 corrective patterns cataloged by Elliott in the 1930’s.  It is remarkable that both these rallies have the same structure and is a testament to the fractal nature of the markets and possibly that the underlying “stimulus” in the former case was provided by the housing and consumption boom (i.e. the consumer and corporate overleveraging) and in the latter case the “stimulus” for the rally was provided by governmental borrowing and spending (i.e. governmental overleveraging).

Figure 4

Other Elliott Wave concepts found in literature:  There is a lot more depth to the Elliott Wave Principle, including the catalog of the 2 motive and 11 corrective patterns and their characteristics and substructure, the entire Fibonacci basis for Elliott Waves and how key Fibonacci ratios exhibit themselves in the Elliott Wave price and time structure and in the individual patterns which provides a useful guide for setting price targets and probable trajectories for future price action, observations on how motive and corrective waves form within channels, etc.  These can be found in an online free tutorial at Elliott Wave International (EWI), a company founded by Bob Prechter that promotes Elliott Wave analysis.  The link is http://www.elliottwave.com/tutorial/ and requires signing up for Club EWI which provides marketing updates and free information releases from EWI.
Proprietary development:  One of the most common criticisms of Elliott Wave analysis is subjectivity, with a standing joke being that a group of 10 Elliotticians would come up with 10 different wave counts for the same price chart.  We have found the Wave Principle to be robust, however, and that the markets always adhere to it, even though an analyst may be wrong about the wave count a market is on and/or a projection or forecast.  We have developed a unique approach which significantly increases the objectivity with which a historical wave count and hence probable future directions can be determined for an asset class at all pertinent degree of trend.  This approach utilizes three key methods to arrive at an objective wave count for the market in question.
Price Momentum and Investor Sentiment:  As mentioned earlier, each wave (or sub-wave, or sub-sub-wave) has a unique nature that is reflected in technical and macro-economic indicators.  We have studied and catalogued the characteristics of the wave sub-structure in terms of objective co-incident criteria in specific technical indicators for price momentum like the ROC, the RSI, and Jurik and Custom Oscillators, which helps us identify the specific wave numbering as well as eliminate possible wave counts for the Elliott Wave fractal fit for the asset class.  We utilize sentiment information like put-to-call ratios, investor and advisor sentiment, etc. to corroborate Elliott based analysis that is reflecting an intermediate term peak for the asset class.  We also have a dashboard of selected technical indicators based on price momentum, breadth, and volume which we use to corroborate and confirm the results of the Elliott based analysis for the asset class especially after the asset undergoes a trend change.
Inter-market analysis:  We have found many market indices and asset classes to be correlated with each other in fascinating ways at a co-incident structure or sub-structure level, or with a lag, which can provide useful clues in eliminating possible wave structures for the parti-cular market or asset class and highlights the most probable wave counts and projections.  Sometimes, these correlations arrive and then disappear.  In addition there are many occasions, especially since the aftermath of the 2000 internet bubble, where liquidity flows through the system raising and lowering entire asset classes with a uniform lag between the corresponding peaks/troughs between the asset classes.  One specific example is that it is common knowledge that the major domestic stock market indices are correlated with each other.  What is not well known is that the Elliott Wave structure and sub-structure for the Dow, S&P 500, Nasdaq, and the Russell 2000 indices has been in lock-step since 2000.  Another example is that there have been periods of time (multiple years in some instances) when the domestic indices have been strongly correlated with currency pairs, like the JPY/EUR pair in the period from 2004 to 2008 and the EUR/USD pair in 2009 as highlighted in our January 2010 market letter and shown in Figures 5 and 6 below, respectively.

Figure 5

Figure 6

Macro-economic Overlay:  Macro-economic indicators typically lag our trend-forecasting models at the intermediate-term.  However when interpreted correctly, global macro-economic indicators provide a unique and objective insight into the most fundamental question – Whether the secular or long-term trend for the local or global economies is inflationary, deflationary, or dis-inflationary.  We monitor a unique mix of indicators to help answer this question which in turns provides an overlay and guide for the Elliott Wave and technical analysis and also an investment bias in our Dynamic Macro series of Absolute Return portfolios.
Successful Elliott Wave forecasts:  Even though the market is always on an Elliott Wave per our understanding of the nature and form of the markets, it takes considerable skill on the part of the analyst to discern the context and to project and forecast price action.  There are some remarkable forecasts based on this methodology in literature as well as in our personal experience at THE ABSOLUTE RETURN, of which we will document a few below
Hamilton Bolton’s forecast for Dow 1000 in 1957:  A. Hamilton Bolton, in the 1957 Elliott Wave supplement to the Bank Credit Analyst, when the Dow was 450 forecast that wave III in the Dow would take the Dow to 1000 in the early 1960s.  In 1960 he gave two price targets, 777 and 999 for the same wave (see Figure 7 below).  He re-iterated the forecast for 999 in 1964 when the Dow was past 777.  The Dow peaked in 1966 with a closing high on February 9th, 1966 of 995.82 (intraday high was 1001.11).  After that the Dow did not break through the 1000 barrier conclusively until 1982.

Figure 7

Robert Prechter’s forecast for the Dow bull market in the 1980s:  Robert Prechter was one of very few analysts in the early 80s to forecast a massive bull market for that decade.  The following forecast and chart is excerpted from Prechter and Frost in The Elliott Wave Principle.

Figure 8

The 1987 Dow peak was 2722.  Compare that with the chart above.  Prechter is also famous for predicting the crash of 1987 two weeks before it occurred.  Prechter got bearish in the 1990s, not expected the Wave V to extend to 2000 like it did.
Our forecast for the Dow/Nasdaq bear and bull markets in 2009:  In our January 2009 newsletter, we made the following specific forecasts (Dow was at 8281 on 1/16/09 when the newsletter was published) based on the Elliott Wave Principle.
      1. Dow would first drop 15%+ below the November 2008 low of 7400 – Dow bottomed on March 6th, 2009 at 6469.
      2. Dow would then rise dramatically in 6 to 12 months to above 10,000 – Dow ended 2009 at 10,428 and subsequently peaked on April 26th, 2011 at 11,268.
      3. Forecast the Nasdaq 100 index to rise to 1700 in 2009 – The Nasdaq 100 (which closed at 1198 on January 16th, 2009, the date of publishing the newsletter) rose to 1860 by the end of 2009.
At THE ABSOLUTE RETURN, we use this objective version of the Elliott Wave Principle as a cornerstone of our technical forecasting methodology for determining the following:
  1. Context for price action for an asset, i.e. what kind of trend prices are in currently (secular, primary, intermediate, short-term) and where are they in the context of the overall historical wave structure for the asset.
  2. Whether prices are moving in an impulsive or corrective manner, i.e. whether a rally is real or is liable to be retraced completely by a future move
  3. Likely levels prices could retrace to during corrections, and targets that prices could get drawn to in motive or impulsive moves
  4. Likely price direction and trajectory
You can get a glimpse of our current forecasts based on the Wave Principle and other proprietary forecasting tools at THE ABSOLUTE RETURN Market Forecast Blog.

Comments are closed.