That weekly candle of the 11 week cycle is getting greener all the time. One clue to the S&P is one of it's componants. The chart is the S&P overlayed with GE and you'll see it bottoms about 20-30 days before the S&P. Of course the market could do the unexpected, but a HUGE push up to December 1-10 time frame is very possible.
I had a whole page of squares and square roots made, only to see at the last second I had one price wrong so it was all wrong. Too much Halloween last night. Are you long or short? I'm short as of Friday , until the market proves itself above 1054. I have a cycle date for Nov 8, which I will look again to see if it's OK to go long again. The chart is a 77 week cycle, and it's pattern is low-low-high. The 97 low, 98 low and 2000 high. Then the 2001 low and past March low, with a high coming. The half cycles are for cycle research, to show a left and right translated cycle. This move has the look of a right translated cycle and so a late ending would work, like 1552 days from March 2000, or sometime in June of 2004 before a crash like decline in late August 2004. But the high for the year is where I'm at now, with Dec 11 as the date , and either 1093 or 1115 as the price. The price can come sooner and Dec 11 will be the test of the high. This all depends on what happens between now and Nov 8. I've been waiting for Oct 30 and Dec 11 since the March 12 lows to start buying puts so now I just have to listen to what the market whispers to me.
I'm Flat on futures closing the week with strong profits; waiting to buy it low since he well reacted down to the ellipse at 1051; any point consolidating above1052/1054 area is - IMHO-a good long scaling out over the week.
I'm long on some stocks geometrically good and with an acceptable risk/reward ratio ( already in profit)
I'm in with options on S&P with a Bear put strategy ATM (Dec 1040 & 1035) (little profit)
I'm Short on Beans ( Losses)
I will go short on gold Monday according to the Stocks behave.
PS. What is the DUE meter you have below the SQ9 of Global?
I do not have it on mine.
Playing with the connections of March 24 2000 to Sept 21 2001 and then to March 12 2003 and Dec 11 2003. DU Meter is @ the link http://www.dumeter.com/scrshots.php
Some number stuff and seperate Square of 9 stuff adding the days from March 12 lows and finding cycle low dates.
low-to-low or high-to-high dates and subtract the number of days or weeks
and then add the square roots of numbers to them.Add the result to the start
date .I guess you could say a generic growth cycle.
Goes something like this, 11 weeks times sqrt 2 is 15.5, add 15 weeks to July 24 and get Nov 6. Or round up 15.5 to 16 and get Nov 13.
Either way it's a cycle top or bottom.
July 24 2002 to Oct 10 2002
78 days or 11 weeks this table uses 11 weeks
but as I pointed out last week , I am convinced of a BRADLEY INVERTED . So IMHO taking in account the yestreday Lunar eclipse, I will be whatching carefully limiting my operativity until the end of the week . In this chart
I humbly foreseen a 1064. We hit a 1060. By next days I - always humbly - believe that the thrust isn't over . Not a 1200 but a good thrust. and then finally down.
this is my humble opinion.
I admit that the market going into year end will be very interesting, but what if at the March 12 lows of 789 I said the S&P would touch 1050. Would you think I was crazy then? You would say I needed my head examinied but look where it is now. The Bradley model has many flaws that software has improved upon. Even Bradley's original presentation was not exact but he waited for the right time so it would be best curve fitted. This first chart is what he presented.
Comment