It’s Alive!

by gmak

I was surprised today to see this, especially since it made me question my own work (which is a very good thing – believe me):

http://slopeofhope.com/2010/03/fujisan-and-the-gartley-pattern.html

This suggests that the Gartley pattern is still valid. So, I did some more reading on the Gartley proportions.  Long story short – I thought I messed up in my computations. I found this explanation here (which I liked because it has some time proportions which help lay out the pattern based on the initial X-A segment.

http://www.fx360.com/techAnalysis/chartPatterns/bearish/gartley-pattern.aspx

The bearish Gartley Pattern Rules

I found this which suggests that there are many possibilities for the ratios. It is from a bullish pattern, but I would imagine that these proportions could be applied to the bearish pattern (for a topping reversal) as well. This suggests that

http://www.harmonictiming.com/free_articles/gartleyarticle.html

Characteristics of the “Gartley 222” Reversal Pattern

1. The swing from Point A will terminate at Point D. This will be at the 0.618 or 0.786 retracement level of Swing XA 75% of the time. The other 25% of the time the retracements will be 0.362, 0.50, or 0.707. A 0.50 retracement is a strong pattern. There have been books and trading systems written about and designed around 0.50 divisions of swings in time and price.
2. There “should” be an CD = AB pattern observed in the move from A to D. At times CD may equal 0.618, 0.707, or 0.786 of AB. At times CD may equal 1.272 of AB.  
3.  The BC move will typically be 0.618 or 0.786 of AB. In strongly trending markets expect only a 0.382 or 0.50 retracement.
4. Analyze the time frames from Point X to A and A to D. These time frames will also be in ratio and proportion. For example, assume that the number of time bars from Point X to A is equal to 17 bars and the time bars from A to D is equal to 11. Eleven is approximately 0.618 of seventeen.
5. There will be a few instances where the CD = AB move will give a price objective at Point X. This will be a true double bottom or double top formation.
6. If Point X is exceeded the trend will continue to move to at least 1.272 or 1.618 of the X to A move.

And I looked here:

http://www.investopedia.com/articles/trading/05/040405.asp

I cannot see anywhere where the BC segment is a shallow one. It seems that it should fall 61.8% of the AB rising segment (in a bearish pattern). Am I missing something or is the Fujisan drawing and the Slopes overlay of SPY valid?

The diagram seems clear to me. Point B should be 38% of the XA segment length added to the price at point A, or 62% of the XA distance subtacted from the price at point X. Point C should be 38% of the AB segment price change added to point A, or 62% of the AB segment subracted from point B. At least that is how I interpret it.

Given the above, I don’t see how this can be a valid Gartley pattern due to the fact that the BC leg did not get deep enough .  The second point is that AB is 61% (8 bars) of the 13 bars from X to A. The above Gartley information suggests that this would be the number of bars for A-D.

Although I really, really want it to be a valid Gartley pattern – wouldn’t that make life easy – there seems to be too many strikes against it being so, in the shallowness of the BC wave, in the time length of the implied A-D move.

X = 1150.45

A = 1044.50

B = 1112.42

C = 1086.02

My original Gartley computations were:

Here are the numbers I get for SPX:

X = 1150.45 (Jan 19)
A = 1044.50 (Feb 5)
B = 1109.98 (Feb 19, 22) – Perhaps one can make allowances for 3 points divergence of reality from the prediction of the pattern – especially that some interpretations suggest different possible lengths of some of the segments.
C = 1058.50 – 1069.50 (projected range) –This is where reality differs from what a Gartley pattern should be.
D = 1124 – 1135 (projected range) Go short here if the pattern holds

If any of the smart people who frequent this blog would care to provide their input on this, it may help us all resolve probable resistance and support levels for a turn. In the end, it may turn out that there is no fool like an old fool – if it turns out that I have misinterpreted something critical.

Thanks.