Full Version: Fibonacci technique explained with charts
Net
Starting February 3rd, I posted Fibonacci expansion targets as a way of giving back to the group. I use a spreadsheet to calculate the targets, and pull the data from recent wave highs and lows. Therefore, the posts do not have charts. Since charts were requested (and there have been questions about the technique) hopefully this post will provide clarity to those questions.

Retracements are pretty common, and almost all charting software draws retracement lines on the chart.

Expansion targets are perhaps a bit less common. Fibonacci expansion targets require a wave and a correction of that wave, and assume the following (projected) wave will conclude in the same direction of the first wave. They require 3 points to calculate: A completed wave high and low, and a correction of the wave from which the point is taken for the expansion projection. If the expansion targets are not realized, then the wave count is brought into question.

The Fibonacci retracement targets require the correct identification of the current trend. Keep in mind that trend is "time frame dependent." The yearly, monthly, weekly, daily, and hourly trends can all be in conflict. Therefore, it is important to identify time frame when identifying the trend, particularly when using the retracement levels.

Since the trend in different time frames can be in conflict, it's useful to look at targets from different time frames and different directions. I.e., an expansion level from a shorter time frame can find confluence with a retracement in a longer time frame (or visa versa). In addition, multiple retracements (corrections) from different time frames can also add confluence, for instance, when a 61.8% retracement on a shorter time frame closely matches a 38.2% retracement on a longer time frame. The same is true with expansion targets (impulses) from different time frames. These confluence areas support each other, and give the targets increased credibility. When there are other supporting areas on the charts, i.e., moving averages and congestion areas, then the credibility of the target becomes significant.

Looking at both techniques, you could say for every A-B-C move, after the initial “A” move: the “B” move is projected using a retracement and the “C” move is projected with an expansion.

Review of Projections:

With this basis, I've charted some of the previous projections showing price data up to the date of the post, and marked the targets on the chart. Each subsequent chart shows the subsequent price action with the next set of targets.

As I stepped through the posts, I tried to leave the previous analysis on the subsequent charts, but this required redrawing some of the lines because of the technique I used to produce the charts.

February 3rd chart:

Post:

The chart identifies the wave from point A to B, and the correction of that wave at point C. The expansion targets assume a move in the direction of the first wave, i.e., a continued move to the down side. The next projection (February 9th) was posted as soon as it was clear that the previous wave was completed.

February 9th chart:

Post:

The Fibonacci targets ended up being accurate with the confluence of support found on the chart. This implies the wave count is correct. Based on the wave count, the next target is identified on the Feb 9th chart.

Ten days later, it appeared that the B wave correction was completed. The next post suggested that the correction would resume, and C wave targets on a larger time frame were posted.

February 19th chart:

Post:

Since the projected target was accurate, it implies that the wave count is still correct. Based on the wave count, and now with more information available, it becomes possible to project from a larger time frame for the conclusion of the C wave. The C wave target remains in play until the wave count is invalidated. If the C wave target is realized, then the question will be if the correction is over, or if the C wave ends up turning into the 3rd wave of a bear run. Targets can be determined at that time to help answer the question.

And, finally, a look at the current chart:

Posts:

Since the post of the 26th, the NDX and COMP pushed up to marginal new highs, representing a test of the 50% retracement area, before the markets sold off. The markets bounced off the lows of the day, so looking at an Elliott wave possibility, here’s where we might be:

With the assumption that the impulsive wave one of “C” completed, we are now correcting that first wave down. This 2nd wave correction could be an “abc” pattern (I'm calling this “little abc”). Using this assumption, we’ve completed “little a” up, and “little b” down. If this is correct, then “little c” to the up side should complete before the 3rd wave of C commences to the down side. Based on this assumption, here are the Fibonacci expansion targets for “little c” taken from the most recent wave up and correction on Friday:

Little c upside expansion targets:

COMP:

0.618 = 2051.8
1.000 = 2072.2
1.618 = 2105.2

NDX

0.618 = 1480.23
1.000 = 1492.80
1.618 = 1513.13

SMH

0.618 = \$41.93
1.000 = \$42.69
1.618 = \$43.91

Looking for confluence between the longer time frame projections and the shorter time frame projections, it appears the little c 1.000 expansion targets are the most likely. These numbers can be compared to the previous forecast from February 23rd, (a larger time frame) which is identified on the current chart for the COMP (link above).

The accuracy of the Fibonacci expansion and retracement targets depends on identifying the correct trend. I use Elliott wave analysis primarily to help me determine the direction of the current wave. I don’t find it necessary to get too far ahead of myself on these projections with swing trading. When the projections are missed, the clear message is the wave count is incorrect. Therefore, there’s value in seeing whether the targets are hit precisely, exceeded, or not reached.

When lower retracement levels hold on corrective waves, this can give a clue to the strength of the subsequent impulsive waves. The post on the 26th considers the possibility that the “little abc” correction might already be over. If so, we will make new lows on short order, this demonstrates substantial weakness--which makes the deeper expansion targets more likely (i.e., the 1.000 or 1.618 expansion targets to the down side).

I should add that Fibonacci levels and Elliott wave analysis works because of mass psychology. Therefore, the techniques are most reliable on indexes where the mass is huge, and will diminish in reliability as the "mass" becomes smaller. There is also the "self-fulfilling prophecy" phenomenon, whereby popular levels are closely watched, increasing the likelihood that the areas will create reactions. (Because of this, I personally rarely trade stocks that have less than 5 million shares daily volume, and most frequently trade SMH.)
Maxwell
Net,

Thanks for posting your thoughts and current EW interpretations. I admit I have problems trying to use this type of analysis, as I tend to see what I'm expecting to see. I appreciate your sharing of your expansion targets.