J.M. Hurst and The 4 Year Presidential Cycle 15

The last time we looked at the NDX and CSCO, we considered the Nov. 2012 low as a Hurst 4.5 year cycle low. In retrospect, this certainly appears to be the case and helps to explain the strong rally in US markets from that low. This implies, in the world of J.M. Hurst, that we have now entered the second 4.5 year cycle of a 9 year cycle period which began in 2009.

In the long view, the 2009 low was likely an 18 and 36 year cycle low. US markets have out performed most world markets and have matched the best runs in history. With 2009 as a 9 year low, we should compare similar cycles early in a long secular wave.  The two charts below show the $SPX in the 70s and the 80s where both 1974 and 1982 were 9 year Hurst cycle lows. The 1974 low was the last 36 year Hurst cycle low and thus the trading out of this major low is worth a look. In fact, both the 1974 – 1982 and 1982 – 1990 nine year cycle patterns are relevant. Both likely witnessed contracted Hurst 4.5 year cycles similar to what we have seen in present day.

SPX 1970s

In the 70s,  the second 4.5 year Hurst low arrived early at 49 months off of the 1974 low. The index rallied from that low to 64 months before a very sharp correction took hold with a low at 65 months (from the 74′ low), taking back most of the prior year’s gains. This was a severe correction relative to the trading at the time.

SPX 1980s

The next 9 year cycle which started in 1982 shows much the same pattern. A contracted 4.5 year Hurst cycle low was seen in Sept. 1986 at 49 months from the 1982 low. The $SPX then rose until the 62nd month from the 1982 low and then witnessed the largest crash since 1929. The bottom of that correction was made at the 62nd month with a retest at the 64th month. So in both 9 year cycles shown here, we have contracted 4.5 year cycle lows from which strong rallies emerged followed by severe corrections. One could argue that these 4.5 year Hurst cycle lows, as midpoint to the 9 year cycle in these examples, were in fact straddle troughs.

SPX Current 9 year cycle

This brings us to present day. The $SPX has rallied from the 2009 low for 58 months in one of the strongest rallies on record. This 9 year Hurst cycle most likely saw a similar contracted 4.5 year cycle low in Nov. 2012 which was 44 months from March 2009. If the same pattern were to emerge here, one would expect the $SPX to top out soon and correct severely into lows this summer, no later than July or August.

While we have similarities in the pattern of trading in these 9 year cycles, there is another strong argument for a serious correction in Q2-Q3 this year and this relates to the 4 year Presidential cycle. Bob Carver characterized this particular cycle in a private forum this month:

“Dewey found a 46-month cycle (3.83-year) cycle which display 98% accuracy over almost a century span of stock indices (he used the Clement-Burgess Index from 1857 to 1883, the Axe-Houghton Index from 1883 to 1952 and another index from 1834 to 1954). He said, These 46-month cycles were observed to occur at 3-4-4-4-4-4 year intervals and then repeat with the 6 cycles recurring each 23 years.

He noted the cycle was M-shaped, with the second peak higher than the first. He attributed the M-shape to two smaller cycles, at one-half the 3.83-year rythm and another at 1/3 (15 1/3 months). He was able to detect all three of these cycles on a periodigram.

Shirk noted in 1960 that the 3.83-year cycle had displayed 98% accuracy over the 127-year period studied to 1960.”

This 4 year Presidential cycle, which runs 46-47 months, all but disappeared in 2005. While the cycle shows clearly on most DOW stocks with lows in 1994, 1998, and 2002, we never saw an appropriate low in 2006 or 2010. One could argue that this cycle was masked by larger cycles at play, and what resulted in the 2008-2009 debacle. Dewey also believed that although a cycle might disappear from view, it would eventually return with the same periodicity and rhythm as before. It is extremely likely that 2014 will see the return of this cycle which is due for a low in July 2014.

So in sum, when trading the next 20 week Hurst cycle, which is likely to start this week, caution would be advised as many stocks are likely to misbehave as we approach a very volatile spring and summer.














About SilentOne

I am a retired Chemical Engineer. I became a house dad back in 2000 and then devoted all my free time to the markets. After years of technical analysis study I decided upon Hurst cycles, which has been my primary method since 2007.

Leave a Comment

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

15 thoughts on “J.M. Hurst and The 4 Year Presidential Cycle

  • William

    I agree that the March 2009 and the November 2012 lows are “4 year cycle” lows. I used the same filter settings Hurst used in Profit Magic to extract the 4 year cycle and ran it on market data from 1870 for the S&P500 and from 1900 for the DOW until today. The current 44 month cycle is almost exactly equal to the 114 year average of the 4 year price wave extracted using Hurst’s bandpass filter. I did the same filter analysis for every wave down to the 10 week wave.

    Bandpass filter analysis dispels with the notion that cycles “disappear”. They are always present. The fact that a particular cycle is not “visually evident” is due to the modulation of its amplitude. It is this amplitude modulation that is also responsible for the variance in a wave’s periodicity.

    One advantage of using bandpass filters for wave extraction is that the amplitude is clearly visible. One can easily see if it is expanding or contracting. For example, the amplitude of the 80 week wave has shrunk considerably compared to its first two oscillations after the 2009 low. Theoretically that may limit the “bounce” out of the upcoming 80 week low unless the amplitude starts expanding again.

    There are other advantages to using bandpass filters for cycle analysis. Hurst has already done the heavy lifting. I highly recommend it.

    • SilentOne Post author

      Hi William,

      Your analysis goes far beyond the Hurst principles taught in the course material. It would be ideal if we could build a bandpass filter tool in Sentient to look at the presence of various cycles. The issue today for Hurst analysts is to see and anticipate a contracted 18 month and 4.5 year cycle in real time. The move out of the Nov. 2012 low was suspect because we did not see technicals nor breadth measures that indicated a 4.5 year cycle low had arrived. You could only make this assumption based on price behaviour and even then it was without complete confidence.

      From the Nov. 2012 low we should see an 18 month cycle that runs close to nominal Hurst values instead of 15 or 16 months (seen off the 2009 lows). I make this assumption merely because the prior 18 month cycle was so short leading into the Nov. low at 13 months. And I expect the majority of the price correction could occur in April with a retest in July. TWT.



      • William

        Hi John,

        One thing I failed to mention in my earlier comment is that I use the average periods from the bandpass filters to create a custom nominal model in ST. I backtested the longer waves for several decades and found that ST could forecast the next two 80 week lows with extreme accuracy. On those rare occasions when the 80 week component modulation was high, the 40 and 10 week waves came to the rescue. The same is true for the shorter waves.

        The fact that we have so much historical data for backtesting and excellent software tools like ST, I have great confidence, from a probabilistic standpoint, in the analysis based purely on price action alone. Even though I’m a Hurstonian at heart, I’m first and foremost a trader striving to reduce my transaction risk to a minimum. ST does a superb job of helping to accomplish that goal.


  • Derek Frazier

    Isnt that the same thing as doing what Hurst suggested and just taking the average of the wave historically and plotting a trough where it should be?

  • Jose


    A bandpass filter tool will defenitely make ST unique, if you know how to fuel ST’s customized nominal model you will be absolutely surprised by the results

    Best for all

  • Derek Frazier

    A test should be done to measure the accuracy of Projection Boxes in ST. Following the guideline of using average periods for troughs when unclear, what is the resultant accuracy for the boxes?

    If ST Projection Boxes are truly calcualting changes in trend correctly, then they should show the approximate moves accurately.
    Although I dont know very much about how the software works, I believe one can make money with it.

    • SilentOne Post author

      Hi Derek,

      In my experience with Hurst and Sentient, everything hinges on the correct phasing. Without this, Sentient’s output will not be reliable and the targets provided by the projection boxes will not be realized. So when you run ST, you must be a good Hurst analyst yourself by setting up ST with proper data and data ranges, pinning the odd cycle low and working the analysis along with ST. That I think produces the best results. The idea of a bandpass filter would take ST to another level as you would be able to study each cycle period (4.5 yr, 18 mth, 40 wk etc.) and fine tune the analysis and projected lows.

      Without something like a bandpass filter, you had to anticipate the contracted 4.5 year Hurst cycle arriving early such as it did in late 2012. A short 4.5 year cycle was likely as we had already seen two 18 month cycles (off the 2009 low) run very short at 16 and 15 months.



      • Derek Frazier

        Thank John,
        I agree. I see that the cycle lows that you are talking about were quite compressed, I also see that when price breaks that 20 week FLD off the lows, and follows through on the weekly chart, traders were rewarded. Going back to the example, if the 2009 lows were compressed and a band-pass filter would isolate the cycles, then when would you have been able to get in? what FLD are you using as an action signal? At the bottom of those lows in 09 I wouldnt take anything less than an 80 day crossing, and probably not before the 20 week. Is there a place to talk with you off this blog that I can pick you ear on how this filter would help a trading strategy like that which David goes with?

        • SilentOne Post author


          Sorry I missed your post. I must have been away at the time. I often post here at this ihub site.

          Stock Charts

          As for Hurst buy signals, it all comes with the right phasing and yes I would go in on a 10 and 20 day FLD crosses. Again it depends on the confidence in the phasing analysis.



  • BobP

    John, I am with you on the March 2009 low being trough of 18y and 9y. My question is the next 54m trough in Nov. 2012 after only 44 months. I’ll need to go back and look at my Hurst cycle course but I thought bull markets usually had extended cycle lengths. I know usually doesn’t always happen so just asking for your thoughts. Thanks, Bob

  • BobP

    William, Very interesting on the bandpass filter. Did a quick search and it looks like there are some Excel add ins for building bandpass filters. As an electrical engineer in my training and original career and also reading Hurst material your analysis is very interesting. Will have to dust off some old mental skills. Did you build your filter in Excel or use some other software? Thanks, Bob

    • William

      Hi BobP,

      I originally did the long term Hurst analysis in Excel. It is relatively easy to do. Next I created user defined functions for various filters in VBA. Those are my workhorses for analysis purposes. Finally I wrote the code in C# for my trading platform.