8 replies to this topic

[airedale88]

GE Hurst phasing of the current 4.5 yr cycle.
this phasing was done using Hurst's fpa tools plus application of a commonality phasing model to resolve any ambiguities. the commonality model was composed of the remaining 29 stocks in the INDU. this commonality model, applied to GE or INDU itself, suggests that the oct 05 low was not a "phase shifted" 40 wk cycle low but a cycle straddle low due to the effect of fundamental events at the time (katrina, refco blow up, etc). we also have a minor cycle straddle at the july 06 low, the commonality 20 wk low having bottomed in june. the muted evidence of the important 80 wk (every 80 wk low is also a 40 and 20 wk low, every 40 wk low is also a 20 wk low)) nest of cycle lows on the INDU and other index charts for the jan/early feb 06 time period shows visual clarity on the GE chart. it is not that evident on index charts due to individual index component stocks showing a spread of 80 wk lows from early jan into feb, cancelling the ability of the indexes to provide an obvious visual low. GE, as are the indexes, is at a 20 wk low this week.

i've drawn in the nominal 80 wk peaks and troughs on the chart (grey lines). the second 80 wk cycle shows a left translated top on GE (indexes show a right transalted top). that suggests that GE's fundamentals deteriorated over that time span. the current right translated 80 wk cycle suggests the fundamentals are improving.



chart....

[airedale88]

just for some more clarity, since many use varied methods to examine cycles, the Hurst nominal 4.5 yr cycle model consists of......... 4.5 yr cycle = 48 to 54 months = three 18 month cycles (nominal 80 wks) = six 9 month cycles (nominal 40 wks) = twelve 20 wk cycles = twentyfour 10 wk cycles etc. each cycle low is a nest of lows containing lows for all smaller cycles.

[mortiz]

Aire, It seems every post you share, I learn another little tidbit about cycles. I find it particularly interesting Hurst cycles do not support the perceived 40 week cycle phase shift that took place in October 2005 using traditional 20/40 week cycle analysis. Phase shifts are not common in traditional cycle analysis (usually once every several years), does Hurst's work ever allow for phase shifts? I realize the Hurst course must be taken to absorb the "refined" version of Hurst's work, but I ran across one of Hurst's early books (copyright 1970, with a price tag of $7.95) and have been finding even his original work is still very close to what I read from you, Echo, and others concerning Hurst cycles. Thanks for the latest update...... as cheif would say, good ********, "that is the coolest thing I have ever seen (famous Butthead quip)", etc. Randy

[airedale88]

randy, thnx. actually the "traditional 20wk/40wk cycle analysis" came from Hurst's cycle course. kennedy gammage and other T/A'ers kind of popularized it but they all seemed to skip over the essential points Hurst made in performing a correct phasing analysis to locate the actual lows. Hurst allows for slight variation in cycle length, but a shift of a 9 month cycle from say 36 to 40 wks avg length down to a 25 wk cycle is too much variation. the Hurst course spends a good bit of time showing how to resolve these ambiguities, something his book does not. there are no phase shifts or cycle inversions in Hurst's work. the course also spends a good deal more time on understanding the effect of both short term and long term fundamental events, both play a major role in creating cycle phasing difficulties.

[chris3403]

aire GE is a dividend paying stock. I am curious as to why you are showing a non dividend paying chart to represent GE’s cycles. I believe we agree 08/02/04 was a 80 wk low and again 79 wks later 02/06/06 was the 2nd 80wk low when one examines the dividend paying chart one sees the top came in on 5/23/05 42 wks after the 1st 80wk low on the chart you are showing it comes in on the 19th wk. Peak to peak is 23wks. According to the principals of Hurst cycles do not invert The first 20wk cycle peaked on its 19th week 23wks later the 2nd peak. From 80wk trough to peak 42 wks, time frame for a 40wk low is supposed to be between weeks 38-42. With all due respect to you and Mr Hurst it sure looks to me like that 1st 40wk cycle inverted. jmo chris

[airedale88]

aire GE is a dividend paying stock. I am curious as to why you are showing a non dividend paying chart to represent GE's cycles.
I believe we agree 08/02/04 was a 80 wk low and again 79 wks later 02/06/06 was the 2nd 80wk low when one examines the dividend paying chart one sees the top came in on 5/23/05 42 wks after the 1st 80wk low on the chart you are showing it comes in on the 19th wk.
Peak to peak is 23wks.
According to the principals of Hurst cycles do not invert
The first 20wk cycle peaked on its 19th week
23wks later the 2nd peak.
From 80wk trough to peak 42 wks, time frame for a 40wk low is supposed to be between weeks 38-42.
With all due respect to you and Mr Hurst it sure looks to me like that 1st 40wk cycle inverted.
jmo
chris

chris, trough to peak measurements are not used in the formal phasing analysis. the top on 5/23/05 is simply a left translated top. it is not that rare for a cycle of any degree to have a sharply (left or right) translated high. adjusting the GE chart for dividends makes no difference in the phasing. you also might have overlooked that i mentioned that GE's phasing involved using a commonality phasing model of the other 29 INDU stocks to help resolve ambiguities in the GE chart.

[Echo]

Chris, not Aire, but i think you have a problem with the May 23rd peak being a peak so close to the nest of lows marking the 40 wk low at end of April 2005. If you look, the April 2005 low was a consolidation move in essence after the sharp rise in 2003-2004. Then the rally off the April 2005 lows failed almost immediately by May 23, 2005 due to whatever fundamental reasons inherent in GE. This can be seen by the left translated 80wk mentioned by Airedale and a break of the VTL of the last three 40wk lows preceding that drop that I've drawn in the same chart. Had GE been destined to keep rising and done so, I don't think you would have had such a problem with April 2005 being a low and no "inversion".

Echo

Edited by Echo, 17 March 2007 - 07:29 PM.

[airedale88]

chris, perhaps a better way to explain the concept if one is not familiar with Hurst's cyclic methods is this. on or around april 18th thru the 29 2005 the majority of stocks in the INDU, SPX, NASD/NDX, RUT, V LINE made their respective nominal 40 wk lows. this low is even visually evident on the majority of stocks and indexes. GE instead shows a narrow range consolidation from march thru april. there isn't much visual evidence of a price decline suggesting a nominal 40 wk low, especially when compared to index charts. that does not mean the 40 wk cycle is not there or it "inverted". it simply means that at that point no selling pressure developed in GE. why? who knows. applying the commonality model to GE indicates that time frame should be a nominal 40 wk cycle low and by listing it as such one can then look forward from that point approx 40 wks +/- to estimate the next 80 and 40 wk low. that 80 and 40 wk low of 2/06/07 came in right on time in the window estimated from using the 40 wk low of late april 05 derived from the commonality model and labeled on the GE chart.

Edited by airedale88, 17 March 2007 - 07:34 PM.

[chris3403]

Echo I’ve been aware of your analytical abilities on this board for a couple of years. However I only know you from this board do you post on other boards using a different alias? Wouldn’t it be a wonderful world if we could trade in real time with the benefit of hindsight? “so close to the nest of lows marking the 40 wk low at end of April 2005” in Apr 05 my friend we were looking for a 10wk low (the 3rd 10wk low) not the 40wk low and the low for GE (10wk) appeared to come in right on schedule during the wk of 4/05 two weeks ahead of the low on the INDU and SPX. Aire I did not overlook the statement “i mentioned that GE's phasing involved using a commonality phasing model of the other 29 INDU stocks to help resolve ambiguities in the GE chart” “that 80 and 40 wk low of 2/06/07 came in right on time in the window estimated from using the 40 wk low of late april 05 derived from the commonality model and labeled on the GE chart” I posted at least 3 times before 2/6/06 that the 80wk low at least for GE was not in even though the 80wk low had been called as being in. Again aire, I’m always appreciative of your comments. thanks, chris

Edited by chris3403, 17 March 2007 - 10:33 PM.