Webinar follow-up


#1

Great ST webinar by David on Monday.

A question came up about measuring the 9 year cycle from the 2000 peak to the 2009 trough and then to the upcoming 2018 peak. The Hurst methodology measures cycles trough to trough or peak to peak not peak to trough. The obvious question is what about the 2000 to 2018 peak to peak 18 year cycle measurement? I noticed the longest cycle in David’s analysis was 9 years. I will post a ST analysis soon on the impact of including the 18 year cycle using the same phasing shown in the webinar. I am curious how this high amplitude cycle would impact the timing of the upcoming peak and composite line. My feeling is it would push the peak forward a bit and make the downturn more severe.

I do not believe that government market manipulation (buying) drives markets to extremes. However, I do believe that ultra-easy monetary policy can drive investor sentiment to extremes. My term for Hurst’s pseudo-trend is bubble. I wouldn’t believe it myself if we haven’t seen it twice before in the last 30 years.

Curt


#2

NASDAQ versus credit spreads. Three six year cycles are shown here. Two nine year cycles were shown in the webinar. Five 42-43 month cycles also can be used. All give the same result with a major high imminent and a major low around 8/2020.


#3

Hi Curt, is that low 8/2020? That is the date I keep seeing as a trough …
The other complication with peaks is that in stock markets you get time translation, and so whereas trough to trough measurements can be fairly consistent peak to peak measurements vary widely.


#4

Curt I like your above “traduction”

I have another one to propose you : “return to reality” means “return to the average”


#5

Hi David,

Yes, 8/2020 is the projected low using a number of different approaches from a number of different starting points. And I agree, highs are not nearly as consistent as lows (sorry for the date typo! corrected).

Curt


#6

First the average. Then below average? TWT.


#7

Pick your poison.

First chart-NASDAQ vs Chicago Fed National Activity Diffusion Index (gives good indications of peaks and troughs in the economy) using Hurst Nominal Model with same phasing as webinar. Projected low 8/2020.

!

Second chart-NASDAQ vs CFNA index using 6 year cycles.Projected low is also 8/2020.

Third chart-NASDAQ vs CFNA using 43 month cycles. Expected low 7/2019 which lines up with 9/2001 low. Extrapolating dates, the 10/2002 low equates to the projected 8/2020 low.

Fourth chart-3m to 10yr Treasury yield curve slope. Yield curve inversions represent peaks in the Fed tightening/business cycle. Not coincidently, the yield curve slope cycle is running at 214 months, same as the stock and credit market cycles above. IMHO, an overly aggressive Fed is what will ultimately kill the stock rally.

Getting the same answer several different ways gives a higher degree of confidence. This also helps me reconcile the confusion in my mind about the Hurst nominal model, the 6 year cycle, and the spectral approach. The ultimate placement of the upcoming peak and projected future cycle trough (and whether this trough will coincide with the price trough) will be determined by cycle dominance.

Curt


#8

The S&P 500 monthly chart below is the 4 year wave based on Hurst’s filter parameters over the last 50 years. I projected the average period of the wave calculated over the most recent 5 oscillations into the future. It troughs in January 2020. The standard deviation of that price wave is almost 22 weeks which creates a large window for the low to occur. There are 16680 days from the 6/30/70 to the 2/29/16 monthly low closes. The average period of the 13 oscillations is 1283 days. Projecting that average into the future from the 2/29/16 monthly low close puts the next low at 9/4/2019. The standard deviation is larger if one identifies the lows based on the low monthly close

It is fairly obvious how to enter as close as possible to the 4 year low. Just buy the 20 week lows inside the standard deviation window because there is a very high probability that one of those lows is going to be the 4 year low.

William


#9

I agree , if Hurst’s 9:18:54:108:216 month model is still valid we should see 9 year cycle low in Sep 2020.
My only reservation, model was not very helpful in the last year opposite to PK WALL model


#11

I also expect trouble over the next 3-4 months.

The projected price level in 2020 seems extreme, but if you look at long term FLDs for support, it is reasonable.


#12

9/4/2019 = 2019-09-04 (sept. 4?)


#13

When I first heard this expression I thought of the shifting that ordinarily occurs near the peaks and troughs of cycles, the distortion that naturally comes from fear/greed causing price to overshoot the mathematical cycle mid-point. I did not think it could apply to a whole market over a longer period of time, but I suppose, when the cycle one is considering is long enough, it is really the same thing?


#14

That is correct. Once the 4 year price wave high is in (projected 3/23/18), the projected wave low becomes slightly more accurate. You can also project the 20 week price wave into the future to see how closely the projected lows synchronize with the 4 year low. Barring any large duration fluctuations they should be fairly close.


#15

The good news is, both Hurst and PK WALL model project significant high in the first quarter of 2018 but then models go a part. I am 60 percent divested.


#16

Hi Scarlet,

From Chapter 5 of the Cycles course:

“A pseudo trend is a fundamental event, rumor, or psychological force, usually favorable, which causes a more gradual increase in volume interest in an equity. This results in a longer term, steady build-up in amplitude of most or all of the waves with proportionally increasing amplitudes. This is a trend-like action in price movement caused only by increasing component wave amplitudes; wave periods, and phases do not change. As volume interest subsides, wave amplitudes also decrease, leading to a steep pseudo trend in the opposite direction, usually downward.”

Maybe it’s just my interpretation, but this sounds like the definition of a blow-off top to a bubble, along with the messy aftermath (in any timeframe). In this case, the event is probably tax cuts (which ironically will likely lead to more aggressive Fed tightening). Importantly, wave periods and phases do not change, so cyclic analysis can still work, although it will be more challenging because of straddled troughs (example-20w trough on 12/29). According to Hurst, volume (and possibly open interest for futures), should give advance warning of a reversal.

Also from the Cycles course, “The skilled cyclic analyst can make good use of a pseudo trend, but the average practitioner is well advised to avoid such situations.” Sounds like “don’t try to top-tick a melt-up”. Disciplined use of Hurst’s action signals should prevent this from happpening. As David said on the webinar, you don’t know whether the high (or low) is in until you get actual confirmation from price at some later date. That doesn’t stop us from projecting it on this forum though😊.

Curt


#17

Hi William,

Do you use the the mid-price or close with your bandpass filters or does it depend on the time period you are evaluating?

Curt


#18

I usually use the close. I’ve experimented with various techniques and they all yield the same proximate result. My favorite though is “digital data spacing” as described by Hurst. It very effectively eliminates the shorter frequencies which may effect the filter output.


#19

Thanks. Is DDS in PM or the course material? I’m not familiar with the term.


#20

It is in Profit Magic. It’s just a fancy term for the selection of your sampling period. Hurst gives some guidelines in PM on the selection of your sampling period to help isolate the wave in which you are interested. Don’t limit yourself to daily, weekly, or monthly times frames. You can create any custom sampling period you want.


#21

Thanks👍🏻. I’ll take a look.