5 replies to this topic

[Rogerdodger]

Only 4 times in 50 years!

Hamzei's blog

Edited by Rogerdodger, 08 March 2007 - 11:29 PM.

[Russ]

It would be a miracle if it did not go to the lower part of the up channel , from the past few months - at about 1300 spx, within the next 5 weeks. These kind of panics historically go down between 10-20%, so far we've seen about 5%.

Low retested within 30 days...
Hamzei's blog

[Russ]

Gotta love this business the next guru under Jason's article - Tim Ord - is looking for spx 1130. "When the McClellan Oscillator gets below -200 a bounce is likely to materialize for the short term. The Summation index touched the -300 level on the current decline. We have labeled in the past where readings below -300 have produced rallies. Resistance lies near the 9300 which equates to the 1442 range on the SPX and may be the area where the next top forms. Next week is option expiration week which usually has a bullish bias and the SPX could rally into late next week. The bigger trend is down and we are expecting an intermediate term decline to take the SPX down to near the 1140 range. Also notice on the chart above as the NYSE made higher highs the Summation Index made lower highs and this negative divergence help pick out the top of March 2005 top, September 2005 top, May 2006 top and the February 2007 top. We are short the SPX at 1381.95. "

[pdx5]

For SPX to drop below 1200, a lot of bad news would have to occur first. Such as housing coninuing its decline thru summer, US $ index declining below 80, oil flirting with $70/barrel, or middle-east heating up. Is it possible? I think so. Is it probable? Not unless atleast one of above occurs.

[A-ha]

Glad to see someone finally noticed that all those past instances occurred after lengthly sell offs unlike this one which came right off the new highs.... However I still find their method of statistical filtering incomplete. They simply use certain numerical thresholds to detect past instances. In my opinion, it is the pattern of the events one should focus on, with more flexible numbers. That way they will be able to extract more valuable information that may suggest a higher probability outcome. For example there is one pattern in this decade significantly similar to what we saw lately. They all miss this specific event in their research because they use hard thresholds for detection.

[Russ]

For example there is one pattern in this decade significantly similar to what we saw lately. They all miss this specific event in their research because they use hard thresholds for detection.

What is it?