The Elliott Wave Theory - Revised Rules and Guidelines

Does Elliott Wave Work?

When it comes to Elliott Wave, there are four primary groups:

1. The Curious: Traders who have heard of Elliott Wave but found it too complex or unreliable.

2. The Devoted: Traders who have mastered Elliott Wave concepts and rely on them in their trading strategies.

3. The Evangelists: Organizations that actively promote Elliott Wave through their financial services.

4. The Pioneers: Companies that have built their business around Elliott Wave, offering specialized software and educational materials.

On the internet, the most vocal group tends to be negative towards wave theories. While this negativity is partly justified, as I will demonstrate, it's important to consider a balanced perspective.

I've often heard Elliott Wave experts claim that the theory becomes increasingly accurate with use. However, my experience suggests otherwise. I spent several years mastering the classic Elliott Wave principles and guidelines, but I no longer rigidly adhere to all of them. A strict adherence would often lead to incorrect predictions.

Elliott Wave is an impressive prototype where wave patterns, structures, phases, moves, proportions, relations, and degrees constantly occur in the market and have a fractal nature. However, they don't normally happen in the idealized ways they were presented.

In fact, I can now demonstrate that the Elliott Wave theory's prediction of 5-3 price movement is often inaccurate. Real-world prices are influenced by multiple contexts, leading to more complex patterns.

As an expert Elliottician, you might think you can label any chart using the three cardinal rules. But the truth is, real-world markets are far more complex. Many charts simply don't fit the simple framework of Elliott Wave, making accurate labeling difficult. Rather than forcing market data into a fixed Elliott Wave model, a more flexible approach is to adapt the model to the unique characteristics of the market.

Does it mean Elliott Wave is useless? Absolutely not! But in order to turn a wave count into a solid trade, one needs to go beyond the prototype and discover other key elements and mechanisms that will consistently work in all trading environments. Some new aspects may be found in the NeoWave and the Harmonic Wave theories, even though adding rigid requirements and complexities is hardly ever the best solution to a problem.

Assuming it takes four years to master Elliott Wave analysis, students typically progress through four stages:

1 - Fascination (0-2 years) - a feeling of comprehension of price action, reinforced by popular study materials that happen to present mostly the idealized Elliott Wave structures (uncommon in real markets).

2 - Realization (2-3 years) - recognizing the limitations of the tool because too many charts don't really abide by the established rules and guidelines. The more you know, the more possible patterns you can build from incomplete price formations, which can create more confusion or frustration.

3 - Contemplation (3-4 years) - trying to incorporate alternative or novel elements to the wave principle that would help to minimize the number of possible wave pattern scenarios and better explain market behavior.

4 - Decision (4+ years) - if incorporation of the new elements was successful (it could take another couple of years or more), the analyst may achieve a new level of understanding price action. Otherwise, the remaining options are:

• Abandon the wave theory entirely,
• Restrict its use to select equities that closely align with idealized Elliott wave patterns,
• Continue using the wave theory by strictly adhering to its rules, even if it leads to forced interpretations of most charts. This approach mirrors the limitations of automated Elliott Wave detection software.

If you don't want to invest years mastering Elliott Wave or its derivatives, and recognize the limitations of its predictive power, I suggest focusing on the core rules and functional guidelines. This foundational knowledge can significantly enhance your analytical skills and facilitate communication within the technical analysis community.

On the other hand, if you're willing to commit years of dedicated study and experimentation, the wave theory can become a powerful tool in your arsenal. It can help you develop a deep understanding of market psychology and identify hidden patterns that other analysts may overlook.

One challenge in verifying progress in Elliott Wave analysis is the potential for stagnation. If you are labelling waves in the same fashion you did when you started feeling confident in your understanding of price action, your progress may be limited. Elliottians must adapt to market changes.

REVISED ELLIOTT WAVE RULES

FOOTNOTES FOR THE RULES

(1) - A new motive wave consists of only three waves:

(2) - The concept of truncation (as well as, to a lesser extent, "expansion" in flats and "running" in triangles) was presumably born in response to the inability to categorize all price action movements using the basic wave patterns. However, it only partially helped because the additions still don't allow to properly explain all the patterns that are present in the markets today.

(3) - As stated in the original rules, wave 2 of an impulse wave can never be a complete triangle (it can only be a correction with a triangle as a subwave B of a flat or a zigzag, subwave C, D, or E of a triangle, or subwave X, Y, or Z of a combination). Nevertheless, patterns with wave 2 as a complete triangle may happen, provided some other conditions are met.

(4) - Wave 3 of an impulse wave is a diagonal. An alternative interpretation would be a WXYXZ correction, but I don't believe they exist (refer to note 11 below):

(5) - On a linear vs a logarithmic scale, the height of the same waves may differ. It may happen that wave 1 of a contracting diagonal is longer than wave 3 on a linear scale, but shorter on a log scale. As a result, the original rule may be confusing and quite often incorrect.

(6) - When wave 4 of a diagonal is a triangle or a flat, then ANY part of the wave 4 correction (not only the last subwave) may enter the price territory of wave 1.

(7) - Wave 4 of a diagonal must overlap wave 1 (it usually happens when wave 4 is a triangle or a combination):

(8) - Wave 4 not overlapping wave 1 only applies to an impulse wave and doesn't apply to a diagonal which should always overlap wave 1. Sporadically, mainly in expanding diagonals, no wave of a correction in wave 4 has to even touch the 2-4 trendline.

Incidentally, there is a lot of misconception, even among CEWA (Certified Elliott Wave Analyst), about where wave 4 in an impulse wave is allowed to travel. The most widely circulated (aka "official") rule is that "wave 4 should never enter the price territory of wave 1." However, the rule should read: "wave 4 should never end in the price territory of wave 1." If you plan to use the classic Elliott Wave, the distinction is important because it can nullify or uphold a developing impulse wave when wave 4 is a flat or a triangle. It is illustrated below:

(9) - When wave B of a flat correction is a contracting triangle, it is acceptable to use wave A of the triangle, as opposed to the end of the correction (ie. wave E), to measure the 80% retracement.

(10) - According to the official Elliott Wave rules, there are three types of triangle corrections: contracting, expanding, and barrier. However, none of these triangles allow for the ACE trendline of the triangle to be horizontal in the bullish sequence ("descending triangle") or the bearish sequence ("ascending triangle"), even though R.N. Elliott himself recognized them in his original work:

(11) - Having scanned and reviewed close to a dozen of WXYXZ combinations labelled by seasoned Elliott Wave analysts (such patterns are unique), I was able to relabel all of them from a WXYXZ correction to either a WXY or an ABC correction without violating any rules or guidelines. In most cases, at least one of the waves originally labelled as wave W, X, Y, or Z was possible to be relabelled as a diagonal or a zigzag correction with a triangle as wave B and/or a diagonal as wave A or wave C. In my opinion, there is no need to add a new type of correction like a WXYXZ (emphasizing that "they are very rare") if price action can be interpreted in a simpler way. Furthermore, just like a motive wave, a WXYXZ correction has 5 primary waves, which directly contradicts the premise that motive waves consist of 5 subwaves and corrective waves consist of 3 subwaves.

Here is an example of a daily S&P500 chart that may be labeled in a complex way (as a WXYXZ correction) or in a simple way (as a zigzag correction where wave A = a leading diagonal, wave B = zigzag, wave C = an impulse wave):

THE RELIABILITY OF ELLIOTT WAVE GUIDELINES

FOOTNOTES FOR THE GUIDELINES

(12) - The guideline of alternation states to expect a difference in expression of similar waves. If a potential bigger complex correction starts out simple at first, then expect complexity to increase during the following parts of the correction (i.e. simple - complex - most complex). The reverse can also apply, but it is more rare. Alternation may also appear in the wave's size, length, duration, extensions, or complexity. Waves tend not to alternate in the corrective subwaves of diagonals and extended waves.

(13) - Illustrations of the Base Channel and the Trend Channel in impulse waves along with the Corrective Channel in zigzags and flats:

The Base Channel and the Trend Channel

The Corrective Channel for Zigzags and Flats

(14) - The Fibonacci relationships are very reliable When a price action pattern is characterized by a close-to-ideal Elliott Wave structures. However, such perfect structures are rare in real markets; the Fibonacci numbers may be more accurate on larger time frames.

(15) - Possible locations of a triangle in a wave 4 correction:

(16) - The apex of a contracting triangle in an impulse wave:

(17) - The guideline of a swift move in the opposite direction after a truncation is extremely reliable if you believe in the idea of truncated waves; otherwise, it has no meaning because the move in the opposite direction is a "normal" start of a new phase. Wave truncations are closely related to the hypotheses of "orthodox tops" and "orthodox bottoms" where the ending points of the pattern don't match the actual price extremes.

(18) - Knowing the current position of a wave within a larger pattern is important because it decides the next fractal move. Partial or insufficient data (especially applicable to IPOs / newly traded securities) can be misleading and risky to trade:

(19) - When working with a line chart, a lot of vital information relevant to price action interpretation is omitted. It is strongly recommended (despite being more difficult to learn) to use either a candlestick or an OHLC chart because line charts ignore the "wicks" (price spikes):

(20) - The preferred wave count is called "primary" and secondary counts are called "alternate." Maintaining both primary and alternate counts may be needed when there is limited price movement information available, which doesn't allow to decipher the primary trend. In all other cases, supporting an alternate count may hinder the whole analysis:

(21) - Comparison of Elliott Wave patterns and their expected positions within a larger formation with popular chart patterns used by technical analysts and traders.

PART 1 - double top, double bottom, triple top, triple bottom, head & shoulders, inverse head & shoulders, bullish flag, bearish flag, megaphone:

PART 2 - rising wedge, falling wedge, triangle / pennant, cup & handle, rounding top, rounding bottom:

About Automated Elliott Wave Software and Tools

Are the modern Elliott Wave tools that auto-generate wave counts worthwhile? In short, no. They just connect the pivot points of the recent price movements that are in line with the rules, without any serious consideration to the guidelines.

The relevant simplicity of the three rules allows for a lot of a flexibility in interpretation of price action. However, the more predictions, the less effectiveness. Moreover, this simplicity cannot be transferred into a competent computer or AI analysis due to nuances of the Elliott Wave theory, such as:

• All, some, or no waves can be extended,
• Extensions can exist within extensions,
• Waves may or may not alternate,
• Waves may or may not truncate,
• The starting point of a wave count may be different from the respective top or bottom of the price,
• A mode of wave development (motive / corrective) may be switched by a single overlap of a wave,
• Volume may be the highest in either wave 3, wave 5, or even wave 1 (in a diagonal),
• Trendlines connecting validation points of diagonals or triangles may cross the structure more than once,
• Breaking a trend channel (via a "throw-over" or a "throw-under") will many times contradict other technical confirmations,
• One subwave of a motive wave can be up to 4 times longer than another subwave,
• It is possible for a wave pattern to break its own rule and still be a valid pattern (eg. a "skewed triangle" where subwave D goes beyond subwave B, or a contracting triangle where subwave E "temporarily" exceeds the extreme of subwave D),
• A completed truncated pattern may be correctly classified as two different patterns on the same time frame,
• A slope of a presumably final wave isn't a dependable indicator of a trend reversal or continuation,
• Corrections and invalidation levels may be as shallow as 16% or as deep as 99%,
• A single correction may be composed of up to three different wave patterns,
• A corrective wave may go beyond the highest / lowest point of the preceding motive wave,
• A final corrective wave may be the same pattern as a non-corrective wave (eg. wave C in a flat correction),
• There are no established time limits for impulse waves, diagonals, or corrective waves,
• The first or the last wave of a contracting diagonal may be longer than all the other elements of the pattern,
• A wave pattern may be validated or invalidated by the chosen type of price scale (linear vs logarithmic) or the kind of chart (line vs bar vs candlestick),
• Invalidation of a single wave or a pattern may prompt a recount of the whole formation,
• It is often necessary to change the time frame to discern waves,
• It is commonly possible to create a zigzag pattern from an impulse wave,
• A continual recount of waves and patterns is required to verify the validity of the primary scenario,
• There is no "wave pattern rank" which would help prioritize the progress of incomplete waves,
• It is virtually impossible to accurately label wave degrees unless all historical data is available,
• A subjective bearish or bullish human bias may distort the whole analysis,
• Mass psychology doesn't work the same as it did when the theory was first discovered; nowadays, it is shaped by dynamic, constant, and multi-dimensional forces,
• Hedge funds and their AI machines have figured out well-known patterns and may use this knowledge, alongside with a bid-ask spread and volume, to manipulate them into new patterns.

The following are four examples of wave labeling that comply with all the Elliott Wave rules, but disregard the guidelines:

The image below shows bad (but very common) examples of wave labels marked by an automated Elliott Wave forecasting and trading software:

One can argue that adding and processing more guidelines on top of modern AI learning systems is going to make the software better. It may be true, but it won't help in forecasting high probability price movements if the established rules and known guidelines are wrong, incomplete, too lenient, or too strict.