Saturday, October 19, 2019

Dow Theory special issue: which indices to use and the appraisal of secondary reactions under the Dow Theory

Answer to an email from a reader

This post is a dedicated response to the email I received from one follower of this blog. I quote the email and my answer in red. To respect privacy I have deleted the name of its author.

I feel the questions were well grounded, and hence readers may learn from the questions and from the answers I give.


The Dow Theorist


As of recent I've been studying Dow theory and have started off by learning the interpretation given in 'Technical analysis of stock trends' by Edwards & Magee. I'm sure you will have likely come across this already.

A couple of points i would like your experienced opinion on:

1.      In the above book (11th edition) it recommends that for Dow theory applied to modern markets it is necessary to analyse the Industrials, S&P 500 and the NASDAQ. Why do you prefer or adhere to using the Industrials, Transports and S&P 500?
The issue of using the Nasdaq (and by implication doing away with the Transports) was also mentioned on the 8th edition of the book you mention. Schannep got to grips with that issue on pages 47, 48  and 98 of his book “The Dow Theory for the 21st Century”. Basically, Schannep says:

1.      The NASDAQ has a small capitalization when compared to the NYSE (and by implication the stocks that make the S&P 500)

2.      The important exchange is the NYSE, not the NASDAQ. Hence, I quote, “[e]very year companies move up to the NYSE, and very rarely do any voluntarily move back to the Nasdaq”.

3.      When the going gets tough, NASDAQ companies tend to plunge much more than other companies. 

4.      Schannep makes a well-grounded defense of the Transports (page 49).

5.      Very importantly, in page 178, Schannep suggest that the proper investment vehicles which are most attuned to his Dow Theory are “American ExchangeTraded funds that track the S&P 500 (Spyders), The Dow Jones Industrial Average (Diamonds) or the New York Stock Exchange composite index (NYC iShares) “ Schannep is not suggesting to invest in Nasdaq 100 shares or ETFs or the like. Apples to apples. We derive signals with some indices and we invest with the indices that are most attuned to the signals given. This is important

6.      Schannep shows the bottom line: By using three indices (INDU, TRAN and S&P 500) results have been outstanding, in fact, much better than those of the "Classical" Dow Theory. I personally dissected Schannep’s performance in this post:

Since I don’t want to plagiarize and quote several pages, I encourage readers to read Schannep’s book in order to be convinced ( ).Read this book until it becomes dog-eared. No pain, no gain.

2.      The above books interpretation also, unlike schanneps dow theory, identifies the confirmation of primary and secondary intermediate trends with a criteria of  1. Being at least 3 weeks in duration and 2. Retraces a minimum of a third of the previous intermediate trend. I have Schanneps dow theory book here and am immediately going to move onto reading it after finishing studying the above mentioned interpretation but having glossed over schanneps dow theory, 3% to identify primary or secondary trends seems quite small a move in the fast moving modern markets (even for averages). Have you found it to be more than feasible in your own practice? I would imagine that the primary trend signals would be quite erratic applying that method.

I guess you are referring to the definition of a secondary reaction (which under Dow Theory perspective would be the secondary trend). The retracement of a minimum of 1/3 of the previous intermediate trend (I’d prefer to call it following Rhea “bull swing” which is the price advance from the lows of the last recorded secondary reaction to the high previous to the new secondary reaction) is a tenet of the classical Dow Theory which is echoed by other authors. 

As to the time requirement Rhea talks of three weeks too, albeit he also makes clear that this is not carved in stone (page 61 of his book, “The Dow Theory”, Fraser Edition 1993). Schannep reminds us (page 18 and 103) that Hamilton (Dow’s understudy) shortened the time requirement by merely demanding “a few days to many weeks” from which Schannep also accepts secondary reactions that only last 2 weeks.

On the other hand, Martin J. Pring (Technical Analysis Explained, Fourth Edition, page 39) advocates in favor of using at least 4 weeks to define a secondary reaction. Hence, as presciently, Rhea wrote (page 61 of this book), “no two students would agree on any rule for selecting and tabulating the important secondary reactions which have occurred”,

As I wrote recently, the more the Dow Theory becomes second nature to me, the more I am convinced that extent and time interact.

In other words, if I have a big retracement (i.e. 50% on a confirmed basis) I maybe should not be so strict when demanding three weeks. Conversely, if I have a mild retracement (i.e. 29%) but the decline has lasted more than three weeks, maybe I should be more accommodative as far as the retracement requirement is concerned.

The good thing of Schannep’s Dow Theory is that is gets rid of such grey zones which ail the classical Dow Theory (which by definition lends itself to interpretations, which will result in some good Dow Theorists, as Rhea was, excelling, and others failing). Schannep provides hard and fast rules which is a boon for the average investor. For sure Rhea could have done away with Schannep’s Dow Theory as he himself really had the instincts of a successful trader (or maybe not, and he would have embraced Schannep’s Dow Theory, as Rhea was convinced that the Dow Theory would be bettered by the passing of time, a he writes on page 5 of his book) .

Yes, Schannep does away with the 1/3 requirement. For him the extent requirement is a confirmed decline of at least 3%. While it may look as too little, Schannep’s track record (which I dissected here) proves that his Dow Theory does not lose the ability to separate signal from the noise when compared with the classical Dow Theory. If greatly outperforms the classical Dow Theory (ca. 200 basis points of additional annual performance and drawdown reduction).

Furthermore, bear in mind that Schannep and I have tabulated the average secondary reaction. Almost none of them has a decline of just 3%, being the average (I don’t have my spreadsheet at hand, so it is not an exact figure) more in the vicinity of 7%. Hence, primary bull and bear market signals determined by using Schannep’s definition of secondary reaction (8 trading days of declining/rallying prices as average of the three indices with at least 2 calendar weeks on two indices and a confirmed decline/rally of at least 3%) has resulted in even more precise signals than those given by the classical Dow Theory. I explained in depth this in the link I have given above.

Lastly, thank you for running this blog. It really is a great resource and even in selecting reading materials it helps to point you in the right direction. I chose schanneps book as part of my Dow theory study because shortly after finding your blog i found your review on it on amazon. You are very welcome.



  1. In your reader's Point #1 he indicates that in the 11th edition of Edwards and Magee's book the Dow theory should be based on the movement of 3 averages: the Industrials, the S&P 500 and the Nasdaq.
    This judgement was made by the current editor of that book, Mr. Bassetti, and not by Edwards and Magee who held to the original Dow Theory which took into account only the Industrials and the Transportation indices.

    Hal Goldberg

  2. Your reader who cited the 11th edition of Edwards and Magee's book concerning the use of the Industrials , The S&P 500 and the Nasdaq as the indices for interpreting the Dow Theory was not indicating the original point of view of Edwards and Magee who held to the original Dow Theory's use of only the Industrials and the Transportation indices. The use of the aforementioned use of 3 indices was written by the current editors of the book, and not by Edwards and Magee.

    1. Good comment! And a very important one! Furthermore, I'd like to add that those that come with new ideas should back them with performance (and drawdown reduction). If the Nasdaq is so good, then, authors of the book, please provide readers with theoretical results by using the Nasdaq as an index since 1985 (I think this was the date of inception of Nasdaq)and compare them with either Schannep's or classical Dow Theory