9 replies to this topic

[SilentOne]

$SPX Hurst Analysis

Been posting at ihub occasionally, but otherwise my market days have been light the last year.

http://investorshub....age_id=79016026

Time for an SPX Hurst analysis update. We did see a 40 week Hurst cycle low in June. We are in the very late stages of the 4.5 year Hurst cycle. An important bottom is due late this year.

I will give a full Sentient Trader analysis in the next post. What I would like to show is the possible path from here. A top is likely in for the current 20 week cycle off the June lows. I have a potential sell for a decline into a 20 week cycle low due mid-late Sept. From there another brief rally takes place into October and may even stretch to the US elections. But from there we fall into a major nest of lows late this year.



The break of the weekly CCI (4) trendline shown above is an early warning for a top to this 20 week cycle.



I would be looking to add short positions late this week (ie. into the JH meeting) for a ST trade. The larger decline will only likely take place after the US elections.

cheers,

john

P.S. I hope Doc can chime in.

Edited by SilentOne, 29 August 2012 - 12:52 PM.

[arbman]

I thought the 19 week samples for nominal 20 week cycle is due around the middle of October since the 35-36 week low in June 5th... I think there is more upside for a week or so and then a month correction, possibly 100 SPX points, if Fed wants to land the equities above 1360 on SPX basis, they'd better push the market to 1450-1460 zone. Otherwise a full retracement is a high possibility. We may just see a stair step down to that lower target after a 20 week cycle bounce anyway...

[NAV]

SilentOne, Your analysis lines up with what i am seeing on the weekly charts. I don't have confirmation yet.

[ogm]

SilentOne,

Your analysis lines up with what i am seeing on the weekly charts. I don't have confirmation yet.

And with the news cycle.

I'm glad we all agree here

[SilentOne]

The current 18 month cycle and the 4.5 year cycle are poised to top out in the coming weeks. The nest of lows for these two cycles is projected for yearend or possibly Jan. 2013.

Here is the 4.5 year cycle view based on 1990 data. Here the 4.5 and 9 year cycle lows were made in 2009. The coming 4.5 year Hurst low is shown that it can stretch out well into 2013. However, based on the contracted cycle periods we've seen off the 2009 low, I think the 4.5 year cycle low will come in around the end of the year. Price should find support at the 4.5 year FLD which is down near the $SPX 1100 -1150 level.



A look at GE confirms the same phasing and similar position with respect to the larger cycles. Note how the various FLDs are forming a uniform upward cascading pattern. Once price breaks the first of these shorter term FLDs, downside price targets start to be generated. So far there is no sell signal.



Here is a view of the SPX current 40 week cycle. As price tops out here in the coming days for the 20 wk cycle, it needs to clear the 40 week FLD as indicated below. Another rise resets the FLDs into a better upward cascade position, and from that point a decline into a 20 week low is set to unfold. Price could find support at the 20 wk FLD mid-late Sept. From there we should see one last rally into the days before or during the US elections. From there we should get a better view of how the correction into a 4.5 year Hurst low is likely to unfold. It could be the largest % price decline we've seen since '09.



cheers,

john

P.S. arbman, I've addressed the various cycle periods - note the consistent contraction off the '09 lows.

sentienttrader.com

Edited by SilentOne, 29 August 2012 - 04:39 PM.

[SilentOne]

Here is the Sentient Trader analysis for the $SPX basis yesterday's close.

The $SPX is in the third 18 month cycle off the 2009 lows. This 18 month Hurst (nominal) cycle has seen 16 and 15 month periods since '09. Cycles have contracted for this cycle and all shorter cycle periods. Average cycle periods are listed in the bottom right corner of the chart below. Note that the cycle lengths for this cycle averaged greater than 18 months prior to the 2009 low. The contraction of cycle periods is very typical of early bull phases when larger cycles have made their lows, such as the 4.5 and 9 year Hurst cycles having bottomed in 2009 (and as discussed further on).

40 wk nominal ~ 33.8 wks
20 wk nominal ~ 15.9 wks
80 day nominal ~ 59.8 days (10 wk cycle)

This first chart is run with ST on its own spectral analysis using data start of Jan 2009.

The current 18 month cycle and the 4.5 year cycle are poised to top out in the coming weeks. The nest of lows for these two cycles is projected for yearend or possibly Jan. 2013.

Here is the 4.5 year cycle view based on 1990 data. Here the 4.5 and 9 year cycle lows were made in 2009. The coming 4.5 year Hurst low is shown that it can stretch out well into 2013. However, based on the contracted cycle periods we've seen off the 2009 low, I think the 4.5 year cycle low will come in around the end of the year. Price should find support at the 4.5 year FLD which is down near the $SPX 1100 -1150 level.



A look at GE confirms the same phasing and similar position with respect to the larger cycles. Note how the various FLDs are forming a uniform upward cascading pattern. Once price breaks the first of these shorter term FLDs, downside price targets start to be generated. So far there is no sell signal.



Here is a view of the SPX current 40 week cycle. As price tops out here in the coming days for the 20 wk cycle, it needs to clear the 40 week FLD as indicated below. Another rise resets the FLDs into a better upward cascade position, and from that point a decline into a 20 week low is set to unfold. Price could find support at the 20 wk FLD mid-late Sept. From there we should see one last rally into the days before or during the US elections. From there we should get a better view of how the correction into a 4.5 year Hurst low is likely to unfold. It could be the largest % price decline we've seen since '09.



cheers,

john

P.S. arbman, I've addressed the various cycle periods - note the consistent contraction off the '09 lows.

sentienttrader.com

[Mr Dev]

hey SO very nice thanks for taking the time to share .

[Echo]

A big hello John! Great to see you posting again. Hope all has settled down with the move and all the rest. Here are my thoughts. On Monday, we will be 13 wks along the 20wk cycle off the Monday June 4th 40wk low. That 40wk cycle was 34.8wks long or 169TD. The previous 80wk cycle ending on Oct 4th was 65.8wks long and was exceedingly short as far as 80wk cycles go. That would imply 40wk cycles of about 33wks in length on average (actual 37 and 29) and 20wk cycles of 16.5wks long. So if we take an average of recent 5 20wk cycles, you would get roughly 16.5 to 17wks in length (unless we start to lengthen again). That targets the very last few days of Sept to the first two-three days of Oct for the 20wk low. There are no significant downside target via FLDs as of yet. But one thing is worth noting. I you use 84TD as the length of the 20wk cycle, then the 20wk fld is 42TD and the 10wk fld is 21TD. The 5wk FLD would be 10-11TD. As of Thursday and Friday, this FLD has been broken to the downside suggesting a downside target of 1385 area SPX cash. Starting this week, the 10wk FLD will be exploding to the upside and within 3 days will be up to the 1401 SPX cash area. So by Wed, we could start to see a break of the 10wk FLD as well. This could trigger a nice downside cascade into the 20wk lows. I would caution everyone however that cyclical price action can be cast aside by strong fundamental news events as will be in play in the next 10 TD. If the market so wants, we could zoom up to new local highs and make the 20wk lows from higher levels. Still, the Thursday-Friday FLD break is the first consequtive 2 day break of this FLD since the June 40wk low (other than an ambiguous break in late July with both price and the FLD falling together). This is an early warning for the bulls that this cycle's top may indeed be upon us. Let's keep the Hurst dialogue going. Its great to see you back. Doc

[SilentOne]

hi Doc,

Glad you posted. I think a decline into Sept. is a good read and bet. It may still take a day or two for that to begin in earnest.

One thing I am looking at is the FLDs further out into Nov-Dec. See how the 18 month and 40 week FLDs are setup? The look may change, but should not, if this is the correct phasing we have. So a decline into a 4.5 year low cannot unfold IMO until we clear the 18 month FLD which takes you well into Nov. Can we rally once again to new highs in that timeframe after the coming 20 week low. My guess is yes. And I will not be convinced of a 4.5 year low until we get a good hard decline that takes out the 18 month FLD to the downside in a substantial manner.



GLGT.

cheers,

john

[Echo]

Hello John, Thanks for the charts. I agree with you that the drop may not start for a few or several days. I see what you as saying on the FLDs and it will be interesting to keep an eye on those FLDs. I "grew up" on Hurst analysis through Airedale who first introduced it to these boards back in 2005. We used barcharts and had a convention of speaking about the cycle length in trading days and included holidays as TDs. One of the key things in Hurst is to have a consistent convention of how the cycles are measured (calendar versus trading days) so that multiple people can have a conversation and are on the same page. One of the reasons I never migrated to ST was time constraints but another big factor is that there is a whole different convention there where the cycles are measured in calendar days and some of the shorter cycles are named as xx day instead of the convention I got used to such as the 2.5wk and the 5wk, etc. So when we talk about FLDs, which are half-span offsets of the cycle in question, what is ST using as the current cycle lengths and what is it using as the offsets. I can then maybe convert them to trading days and see if we are using the same cycle lengths and offsets for the FLDs. I am also not sure how to handle the recent dramatic shortening of cycles in the past 3 years and especially with the last 80wk low in Oct which came at an amazing 29wks for the 40wk cycle off the March 2011 40wk low. Airedale and Hurst recommended averaging the last 3 cycle lengths to get a working cycle length for any particular cycle, but with the out of the ballpark value for some of these cycles, I am more apt to average the last 5 or 6 cycles, right or wrong. I have often felt that shortening of cycles was directly related to available liquidity but never had any proof of that idea. Certainly the world is awash in liquidity right now and that might fit with the very short cycles. I have noted that they may be starting to lengthen again here lately. GL/GT to you too. Doc