EURUSD – addendum 7

I want to add a couple of comments about the long term picture to my prior post about the Euro. The chart below is a monthly chart of the EURUSD with the 18 month price wave and 4 year price wave. Except for the minor frequency modulation in the early 2000’s, the 18 month wave has been very consistent for almost twenty years. The long wave at the bottom of the chart averaged a little over 50 months for over four decades prior to beginning of the so-called financial crisis in 2008. Not a perfect 3:1 harmonic relationship, but close enough. However, beginning in 2008, the very large increase in the amplitude of the 18 month wave has caused the 50 month wave to become somewhat distorted making it more difficult to forecast when the next low will occur. A rough estimate can be made by projecting the filter output into the future at the same rate of curvature. Another observation I would like to point out is the fact that the price action has pushed hard down through the bottom of the dominancy envelope which greatly increases the transaction risk of further shorts. The shorter waves mentioned in the prior post should facilitate getting in close to the 18 month low feasible for the nimble trader.


$EURUSD (Monthly)  8_1972 - 2_2015

About William Randall

I'm a retired attorney who discovered the fascinating world of the financial markets almost two decades ago. After learning about Hurst, I've spent the last several years honing all the skills necessary to fully implement the techniques he taught.

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Your email address will not be published.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>

7 thoughts on “EURUSD – addendum

  • Harman Sidhu

    Hey Mr Randall
    Nice articles on cycles and Hurst nominal model applied to currencies.I am particularly interested in the tools that you use that aid with the analysis ( Dominancy Envelopes and BandPass Filters) could you elaborate on them and the construction process of the envelopes and whether they are available commercially. Your reply would be greatly appreciated.My two cents on the cycles in Euro-Dollar : if we consider the possibility of 80 week cycle inverting after the January Low in 2012 to July 2013 , then the low for major cycles could be already in with January low.

    Thanks & Regards
    Harman Sidhu

    • William Randall

      Hi Harman,

      The bandpass filter design I use is based on the filter design Hurst used in Profit Magic. I doubt if it is commercially available since the formula is available in the book. My envelope technique is proprietary but a search of the internet will provide you with some reasonable options if you don’t possess the coding skills.

      As for the Euro, this week’s price action indicates that January clearly was not the low. I believe the next 20 week low has a very high probability of being the low. I plan to buy it.


      • Stuart Mendel

        Thank you for your analysis of the EURUSD. As as been stated, we are all highly motivated to successfully catch this falling knife utilizing cyclic methods not only for the financial satisfaction but also for the intellectual as well. And so I set out to look at the US dollar index to either support or refute your hypothesis of the next 20 week low in the EURUSD to be “the low.” This would then imply that the 20 week high in the US dollar index would be important.
        As we have mentioned before, Hurst’s nominal model was derived from his spectral model done on the DJIA and other equities. I have not found reference to any extensive work of his on currency pairs nor synchronized peaks v troughs. However i have encountered major disappointment in the application of the out of the box Nominal model to everything and then pinning troughs and forcing expert models to get something to make sense (please don’t take this personally, David). How do I know when I am right (or wrong) before the fact?
        Performing spectral analysis and dominant cycle extraction on an appropriate section of data might be a viable alternate approach. The trick is the data set chosen: too long, and the current information diluted by the older data and too little, unreliable statistical reliability.
        The first chart demonstrates the dominant cycle output for various analysis periods.Going back to 1978, a dominant cycle of 49 months was found. The ideal 49 month plot is shown in white and the actual plot of the cycle shown in red. There is some correlation but not much. But taken from the 2008 lows we see a good correlation with the dominant 35 month cycle. We see a strong correlation of the 18 month as well. Clearly, these cycles violate the usual 3:1 relationship in the Nominal model and 36 months is way off form 54 months. These are the cycles that are present in the data.

        • Stuart Mendel

          So now that I have trashed the Nominal model, I form a custom nominal model, one that incorporates
          1. A 36 month cycle
          2. An 18 month cycle
          3. A 2:1 harmonic ratio
          I find that there is somewhat of a relationship between the time ST takes to do an analysis and its viability. The faster the analysis is performed the more viable I have found it to be. And ST loved this nominal model.
          Cutting to the chase, ST is telling us exactly what we need to further support the premise that the next 20 wk high in the US dollar index will be important and, by extension, the next EURUSD 20 wk low will be “the low.”

    • alain

      First for Hurst there is no “Dominant Cycle ” it is not in his “theory” which for me is not a “theory” but more a methodology.

      Second there is no “Nominal Model” and that is not a “Model” (sorry David)

      The “Nominal Model” comes out of nowhere without justifications, only assumptions from various writers here and elsewhere.

      I mean the “Nominal Model” is not so justified (cycles period based on spectral analysis ?, relation of 2 and sometimes 3 between adjacent cycles etc)

      By the way why “sometimes” we jump from 2 to 3 for the relation between 2 adjacent cycles . And why these particular cycles and not the others.

      And why having discarder the work and research of the FSC (Foundation for the Study of Cycles) of Dewey.

      Dewey and Gertrud Kirk and other researchers found a 6 year cycle as a very effective one in the market (s).

      Why having discarded the 6 Year Cycle.

      6 Year is 2 x 3 Year and ‘4 x 18 months

      Anyway there is no “Nominal Model” but several. At least officially 3 !

      a) the one in Profit Magic
      b) the one in the course
      c) the simplified one in the course for the dummies – you and me with a pencil, a sheet of paper, a rubber and no computer ( 1975)

      by the way no explanations about the difference between the “model” in the book (1970) and the model in the course.

      You just have to take it for granted.

      • Stuart Mendel

        You might want to reread (or read) and understand Profit Magic p.168-index and then reconsider some of these statements.