Short Term Trading Member Forum

[slinky]

Chris, Here's an Airedale post from 6/24/04 (to you in fact) regarding the phasing of GE that explains some important things that Trend and others might find useful. This is not simple stuff and require serious dedication, more than I was ever able to devote:
----------------------------------------------------------------

chris, a cyclic principle not developed in Hurst's book but a key concept in his later research and is used fully in the course is the principle of harmonicity; cycle periods are related by a small whole number, usually 2 except for the 80 wk and 4.5 yr cycles which is 3. combined with the principle of sychronicity that means that a trough of a cycle will also be the trough for all smaller cycles. these are called "nests of cyclic lows". any nominal 10 wk cycle low will also contain the low of the second of the 2 nominal 5 wk cycles contained in the 10 wk. that 5 wk low would also be the low for the 2.5 wk low and all smaller harmonicly related cycles. the 5/17/04 low, approx 7/8 wks from the march 24 low, was by my work a shortened (principle of variation) nominal 10 wk low so therefore it was also a (shortened)5 wk low. we just had another 5 wk low come in and we're moving off that low.
GE may also be in a small pseudo trend (again, from the course, not in the book)due to the rebalancing of index funds.
cycles in the market appear to have shortened since the jan top. this variation may continue but there is no way to tell. a current cycle phasing analysis taking into consideration this variation puts the possible 80 wk low in july, while estimating the time window for the bottom based on much longer weekly cycles which have less samples to average out moves the window into august, possibly sept. tracking the shorter cycles and using harmonicity and synchronicity should help narrowing that window.

aire

mr cash and chris re GE.... labeling the oct 02 low in GE as the last 4-4.5 yr low pretty much breaks every rule Hurst developed for analyzing cycles. here's just a few of them. Hurst's nominal (average length) cyclic model lists a 4.5 yr cycle (54 months), which breaks down into three 18 month {78-80 wks)cycles. should the 4.5 yr cycle, due to variation, come in short as a 4 yr (48 month) cycle, then we would expect three nominal 16 month cycles. using the monthly GE chart shows a clear cut 52 month cycle, from oct 98 to feb 03. that 52 month cycle breaks down harmonicly into the three nominal 18 month cycles, the 1st from oct 98 to feb 00, 16 months, the second from feb 00 to sept 01, 19 months, and the 3rd from sept 01 to feb 03, 17 months. that's 3 cycles of average length 17.3 months, well within the parameters of Hurst's cyclic model. we can also draw a cycle based valid trendline from the nominal 4.5 yr top and it is not broken until after feb 03. here's the link to the monthly GE chart, you can easily see the 3 nominal 18 month cycles and you can draw the trendline if the chart is printed.

http://charts.barchart.com/chart.asp?sym=GE&data=G&date=062404&den=MED&evnt=off&....

we can also observe that GE had only a brief rally off the oct 02 low, not typical of a stock coming off an important 4/4.5 yr cycle bottom. the rally off the feb low certainly is much more typical, it also broke above the 78-80 wk fld which gave a price target right into the highs GE has made so far.

a very important tool which Hurst said must be used when there is any question about the proper location of any cycle is the commonality phasing model, explained only in the course. unfortunately it is not something that can easily be displayed on a chart, at least for someone with the rudimentary computer skills i have. Hurst would take a strip of paper and mark on it all easily confirmed (by phasing methods such as past time flds, vtls, etc) cycle lows from the "stable" of stocks the analyst covers. Hurst's service covered 27 stocks plus 8 or 10 i think commodities. i have created commonality models for monthly, weekly, and daily charts on a group of stocks containing all the INDU stocks, the top 10 by weighting of NDX, the top 10 by weight of the SP500, the top 3 by weight of SOX , the Dow Transport index, NYSE Comp, Value Line Arithmetic index, SP Midcap Index, the BDX index (broker dealer), the Internet Index, and a few others. 50 plus equities and indexes. Hurst explained that the commonality index would show the true cyclic model locations of cycle troughs that can at times be masked by asynchronous fundamental interaction on any one or more cycle(s) in any particular equity. when i completed all three models, even i was amazed by how clearly they were able to remove any ambiguities in so many situations. the monthly and weekly models confirm with out a doubt that the GE oct 02 low was not a 4 - 4.5 yr low. both models also confirm that of all the stocks analyzed, whether INDU, NDX, Midcap, whatever, an overwhelming majority have march 03 as the location of the 4.5 yr low, DESPITE a lower price in oct 02. GE bottoming in Feb 03 was just typical cycle variation. those few stocks out of all analyzed where one could not conclusively label feb/mar 03 as the 4.5 yr low also were those that showed the the widest amount of variation in all the nominal cycle lengths, these were some of the drug stocks and some semis, tho not all, some fitting perfectly.


aire