Well, I've been thinking hard about my Elliott Wave count for the S&P 500 , because it is very difficult to find a clean-looking impulsive wave structure coming down from 2137.1. I originally ascertained a less than ideal count of 3 waves down to 1833.5, but once this upward correction took out the 2070 level this week (violating the extreme of the only wave I could even remotely justify as being a valid wave 1 in this sequence), it made it clear that this upward rally is not a wave 4 (and therefore likely a wave 2), but then I can't find any distinctive lesser-degree 4th or 5th wave within the structure at all.

So, scrapping the whole count since 2137.1, now I have an entirely new perspective on this, which allows for much greater clarity:

Keeping in my view that the year 2000 represented the peak of the 3rd wave of Grand Supercycle degree...the rest has been part of the 4th wave since then.

- Wave "a" occurred from 2000 to 2009.

- Wave "b" occurred from 2009 to 2015.

Type "B" waves are the only other waves beside type "4" waves which can independently develop as triangles (though an "E" wave can be a triangle, it is always the last leg within a larger existing triangle). This is the key point of my change of perspective today.

Looking closely at how this wave "b" peaked at 2137.1, then undulated for 3 months before the index abruptly tanked down to 1833.5, it is possible to make out the pattern of a contracting triangle (waves a,b,c,d,e). It is certainly very late in forming the latter four waves, but it makes an unmistakable pattern. It means that wave "a" of the triangle had been ongoing since the low of 2009 until May 2015, and the date of origin of that wave is where the post-triangle thrust measurement is taken as the distance between the upper and lower converging trendlines of the triangle. At that origin, the measurement is a distance of 800 full points. The vertical cyan line in the chart shows the expected minimum downward thrust from the terminal point of wave "e" at 2107.4 on August 18th. This thrust represents wave "c" vis-a-vis the waves "a" of 2000-2009 and "b" of 2009-2015.

Apparently, the first wave of the thrust began from there and reached down to 1833.5. The three-wave upward (abc running flat) correction since then has been the second wave and so far just exceeded a 78.6% retracement . The second wave should not exceed 2107.4, if I am correct, before the third wave down begins.

In the chart, based on general EW conventions for estimating the relative lengths of subsequent waves based on the first wave's length, I have added presumed levels where waves 3, 4 and 5 might hit. The cyan markers assume a length of wave 3 as 1.618 of wave 1, while the yellow markers assume a 2.618 length of wave 1 for wave 3 -- which interestingly enough seems more in agreement with the projected length of the post-triangle thrust.

Notice also that the post-triangle thrust approximates the level where wave "c" would equal wave "a" (which is a generally accepted EW guideline for typical wave "c" length), which also happens to be near the point that represents a 61.8% correction of the rally from 2009-2015.

It is possible that the extent of this wave "c" is only the end of wave "A" of 4 (wave "A", comprising "a", "b" and "c", having started in 2000) with a rally to come thereafter in "B" and a final crash in "C" potentially bring the index into triple-digit territory in the early 2020s.

check also this one