15 replies to this topic

[airedale88]

this chart is the monthly NYSE Comp. it has the Hurst nominal cyclic model applied over a number of years. IMO the consistency of the model is remarkable. toward the right i have boxed in an area that has caused some difference of opinion as to locations of 9 month cycles to some and an 18 month cycle low to others. Hurst analysis methods using a commonality phasing model indicates mid jan to early feb 06 as the bottom of an 18 month and 9 month low. it is not visually apparent. Hurst discusses anomalies such as this and how to resolve them, the complex elements of price motion can often mute evidence of any degree cycle but that cycle is always present. for Hurst students, selection of any other visually apparent low such as oct 05 or jun/july 06, as a 9 or 18 month low creates a cycle that varies far too much from the average past time lengths. Hurst's methods do allow for variation in cycle length, but it is a slow to change process, not sudden jumps. one simple method to avoid the anomalies of the boxed in area is to extrapolate out from established previous cycle lows. aug 04 was a clearly defined 18 and 9 month low. the end of april 05 is a clearly defined 9 month low. since the 4.5 yr cycle = three nominal 18 month cycles = six nominal 9 month cycles we can count from the aug 04 low two average 18 month cycles to arrive at the early aug 07 4.5 yr low. we can also count three more average length 9 month cycles to arrive at the same point. all of these 18 and 9 month cycles since the march 03 4.5 yr low have shown high right translation (even those of question inside the boxed area). the cyclic trend larger than the cycles of 4.5 yrs and less has to be very bullish to create these right translations for over 4 plus yrs. the only conclusion is higher prices.

[relax]

would you consider using two different cycles one with a 4,5 year cycle low at october 2002, which would explain the bottom of march 2007 and then the one you are using with a low at march 2003, which explains the bottom of augsut 2007 and then when these two cyles nest (december 2007), one could argue for a significant bottom...

[airedale88]

would you consider using two different cycles

one with a 4,5 year cycle low at october 2002, which would explain the bottom of march 2007

and then the one you are using with a low at march 2003, which explains the bottom of augsut 2007

and then when these two cyles nest (december 2007), one could argue for a significant bottom...

no.

[OEXCHAOS]

Could you be more clear?

[NAV]

aire, I am no cycle expert and have no knowledge of hurst formal phasing methods. Correct me if i am wrong. With the assumption that April 05 was the 9 month cycle bottom, and with a presumed extension of the time length, can't the next 9 month cycle bottom be around June 06. That would make the 9 month cycle length about 1.5 times the prior nine month cycles length. If that does not violate your cyclical model rules, then the next 9 month could very well been March 07, giving the timeline for the next 9 month bottom around December 07. Cuz, from e-wave perspective and e-wave time anlaysis, Dec should complete wave C of this entire correction from June 07, which should come around Dec 07. That would also fit nicely with a 9 month cycle bottom. Thoughts ??

Edited by NAV, 12 September 2007 - 07:36 AM.

[LarryT]

Good work, if we combine Fibonacci time extensions to the model and use the March 2003 and August 2004 lows as the time for extensions we get an 18 month low (72 weeks) at 1.618 in time in September 2005, February 2007 at 2.382 extension as nominal times for the low. Therefore the March 2007 low is the low in our opinion at Market Charts/Turns. That stated we can then do a Fibonacci time analysis of the next 72 week low. Using .382, .50 and .618 extensions the top has to occur at 9-21-07, 11-23-07 or 1-18-08. My analysis favors the 11-23-07 at 1605 SP-500 cash index as the top for this next 72 week cycle. Best, Larry

[swingtrader]

Aire, Really nice work. Thanks so much for the update! ST

[airedale88]

aire,

I am no cycle expert and have no knowledge of hurst formal phasing methods. Correct me if i am wrong. With the assumption that April 05 was the 9 month cycle bottom, and with a presumed extension of the time length, can't the next 9 month cycle bottom be around June 06. That would make the 9 month cycle length about 1.5 times the prior nine month cycles length. If that does not violate your cyclical model rules, then the next 9 month could very well been March 07, giving the timeline for the next 9 month bottom around December 07.

Cuz, from e-wave perspective and e-wave time anlaysis, Dec should complete wave C of this entire correction from June 07, which should come around Dec 07. That would also fit nicely with a 9 month cycle bottom.

Thoughts ??

NAV, a cycle that suddenly jumps to 1.5 times it's average length would raise a big red flag in Hurst's cycle phasing analysis. the "missing" 9 month cycle that was due in early 06 can be found by use of a Hurst commonality phasing model. an examination of a set of stocks and sectors noting easily labeled cycle lows. here's three charts i had on file, all suggesting there was a 9 month cycle low in mid/late jan 06.

XLF... 40wks (9 months) from the april 05 low and approx 18 months from the aug 04 18 month low we have a mild pullback after a strong rally and consolidation









BA.... a strong uptrend with mild pullbacks on all cycle lows from march 2003. the late january 06 low is 40 wks (9 months) from the april 05 low and 78 weeks (18 month) from the aug 04 18 month low. note the acceleration up after the jan 06 18 and 9 month low.





GE.....provides a clear cut 18 and 9 month low in the first week of feb 07. a small variation from the late jan lows of some other indices and stocks, and is quite normal. GE also provides clear visual evidence of a 9 month low in late nov 03 which is not well defined on the SPX and other broad indices and a early nov 06 9 month low that just manages to appear on index charts. at the same time, the aug 04 18 month low and the april 05 9 month low are not that obvious on GE as on SPX etc. this is the norm. at any time, visual evidence of any particular cycle low may not be apparent, but using the cyclic principals of commonality and variation (all equities share the same nominal cyclic model subject to minor variation) and/or other hurst tools one can extract the needed cycle info.

[OEXCHAOS]

That's some nice work. Mark

[airedale88]

Good work, if we combine Fibonacci time extensions to the model and use the March 2003 and August 2004 lows as the time for extensions we get an 18 month low (72 weeks) at 1.618 in time in September 2005, February 2007 at 2.382 extension as nominal times for the low. Therefore the March 2007 low is the low in our opinion at Market Charts/Turns.

That stated we can then do a Fibonacci time analysis of the next 72 week low. Using .382, .50 and .618 extensions the top has to occur at 9-21-07, 11-23-07 or 1-18-08.

My analysis favors the 11-23-07 at 1605 SP-500 cash index as the top for this next 72 week cycle.

Best,
Larry

larry, in my experience fib analysis doesn't mix with Hurst cyclic analysis. 2 completely different disiplines. i disagree with all your fib extension lows as indications of the nominal 18 month cycles



regards, aire.