Saturday, January 12, 2013

Dow Theory Special Issue: An Answer to the New Low Observer (NLO)

This post is the consequence of a commentary posted by the New Low Observer (NLO) on an Instablog in Seeking Alpha.

You can find the post and NLO’s comments here:

Since the NLO’s commentary was very extensive, I decided that my reply would be a new post.

I hold the people at the NLO in high regard since they have an encyclopedic knowledge of the markets and, in particular, of the writings of Charles Dow. You can read more about NLO here.
However, when it comes to investing it is difficult to always agree. This is precisely what makes markets work: difference of opinions. Otherwise we would all be buyers or sellers at the same time. NLO’s comment was one of these instances where, keeping a civil tone for which I thank them, there was some disagreement as to the way the Dow Theory is to be applied.

The goal of this post is to analyze the points raised by the NLO. I feel all people interested in the Dow Theory can gain from following this discussion. The more I practice the Dow Theory, the more convinced I am it is full of nuances, and that it is more complicated (but accurate) than it looks on the surface. This post may contribute to highlight the fine print around the Dow Theory.

This is going to be a very long post (something which is discouraged in the blogosphere). Thus, I recommend the readers to read it piecemeal. The upside of it is that it can be very enlightening to us all. As in all walks of life, "no pain, no gain".

I am answering each of the objections raised by the NLO as if I were answering an email. Thus, I have inserted my answers into the text of the original NLO’s comment with a different font and red color to make it easy to distinguish objection and answer. 

Have a wonderful weekend.


The Dow Theorist


Your call on the secondary trend being bearish on October 5, 2012 was FANTASTIC. You hit the exact top in the market at that point in time (found here:

However, as a student of the markets and Dow Theory, I find your January 7, 2012 posting to be perplexing on many levels as it relates to Dow Theory.

First and foremost, You've indicated that although Dow Theory has turned bullish on January 2, 2013, the PRIMARY BULL MARKET trend actually emerged on November 15, 2012. However, when we revisit your November 16, 2012 posting titled “Dow Theory Update For Nov 16: Primary Bear Market For Stocks Signaled Today” (found here: we find that, as your title states, a PRIMARY BEAR MARKET had just begun. Furthermore, you go so far as to say in that article (11/16/2012) that October 5, 2012 was the actual beginning of the PRIMARY BEAR MARKET.


There is no contradiction or mistake. In “real time, one cannot distinguish the first leg of a new bull (bear) market from a secondary reaction against the ongoing primary bearish (bullish) trend.

When a secondary reaction against the existing primary trend develops nobody knows whether such reaction will end up being the “real thing," namely not merely a “reaction” but the first leg of a new primary bull (bear) market, or just this: a humble secondary reaction which does not reverse the primary trend.

On October 5, 2012, in “real time” (not writing 3 months later) I labeled the primary trend as bullish and the secondary trend as bearish. Of course, subsequent events, namely the violation of the secondary reaction lows on mid November, flashed a primary bear market signal and, thus, the correct thing to do under Dow Theory was to re-classify the secondary reaction as the first primary leg of a new primary bear market.

October 5 was not the actual beginning of the primary bear market. The actual beginning was the date when the last confirmed highs were made, namely Sep 14. October 5 was merely an unconfirmed higher high which, being unconfirmed, is to be discarded. By the way, such non confirmation was punctually noticed in my post of October 5.

Thus, in mid November, we ascertained that the market was in a primary bear market since Sep 14.

From the November lows the market staged a powerful rally (bullish secondary reaction against the primary bear market existing since September 14). Then it underwent a pullback and on January 2 the secondary reaction highs were broken out. Such breakout was a text-book primary bull market signal under Dow Theory. Once again, in “real time,"  there was no way of telling that the rally that started in mid November was the “real thing” (a new primary bull market instead of a bullish secondary reaction within a primary bear market) and, accordingly, the first leg of a primary bull market. This could only be known on January 2 and, thus, the proper thing to do under Dow Theory was to reclassify the rally (secondary reaction) as the first leg of the new primary bull market which then started in mid November.

All in all: We had a short bear market that started on Sep 14 and finished in mid November, which became recognizable under Dow Theory in mid November. And we are currently under a bull market that started on mid November, which became recognizable on January 2, 2013.

Of course, neither the Dow Theory, nor any Dow Theorist can be blamed for not being able to call primary top and bottoms in real time. This is beyond the power of the Dow Theory.

My point of view, which is fully orthodox according to the Dow Theory, has been ratified by Richard Russell in his book “The Dow Theory Today” (Fraser Publishing Company, Fourth Printing 2003, page 7). Richard Russell when explaining the transition from a primary bull to a primary bear market says:

“However, if one or both Averages refuse to go over the previous high point (B), and the two Averages then decline (on increasing volume), breaking the low point of the secondary reaction (C and E), we then must say that the bull market has ended and a bear market has been operating since the high point (B). After the bear market signal has been given, we then will go back and reclassify the entire secondary reaction (B-C) as the first primary swing in a new bear market, and the retracement (C-D) will be reclassified as the first secondary reaction (upward) in a bear market (emphasis mine)”


When we review the Dow Theory analysis that followed October 5, 2012, we find that your view was that the market was in a PRIMARY BULL MARKET. This is confirmed by your October 10, 2012 article titled “Dow Theory Update For Oct 10: Secondary Trend For Stocks Turned Finally Bearish” (found here:


Same. The primary trend is deemed to remain in existence until a Dow Theory signal says otherwise. Thus what “in real time” can only be classified as a secondary reaction becomes the first leg of a primary bear (bull) market after the Dow Theory primary bear (bull) market signal has been given.


Your concluding thoughts were:

“Don't press the panic button. The market continues in a primary bull market, and a secondary reaction is something to be welcomed as it is healthy (like a summer storm). Markets cannot go up in a straight line. As to gold and silver, we cannot even talk of a secondary reaction. Don't forget that the action of October 4th, was very bullish for gold and silver. Remember what I wrote in this post.”


Bearing in mind the preceding clarifications, clearly I was right on October 10, 2012 to label the primary trend as bullish and declare that the decline hitherto experienced was merely a secondary reaction. Of course, once we got the primary bear market signal in mid November 2012, and strictly according to the Dow Theory (even with the most conservative “flavor”), the only correct thing to do was to reclassify the downward swing  as the first leg of the primary bear market.


Again, let’s be clear, your call on the secondary trend being bearish on October 5, 2012 was FANTASTIC. You hit the exact top in the market at that point in time.

However, it would be disingenuous to “RETROCAST” your analysis by pointing to when the trend changed and then overlay Dow Theory on top of that and then say, “see Dow Theory was right.”.


I was not “retrocasting” but merely re-classifying as new market action becomes available. Until Mid November 2012, the downward movement could only be classified as a secondary reaction against the primary bullish trend. After the primary bear market signal of mid November 2012, then we had to classify such a downward movement (started on Sep 14) as the first leg of a primary bear market. There is no way of knowing in real time, whether the downward (upward) movement is merely a secondary reaction within a primary bull (bear) market, or it is the first leg of a new primary bear (bull) market.


It is an injustice to Dow Theory which is intended (based on the Hamilton/Rhea version) to track PRIMARY BULL MARKETS which tends to last from 2-4 years and PRIMARY BEAR MARKETS which usually last from one-third to three-fourths the preceding PRIMARY BULL MARKET.


Based on the above, I feel I am not doing any injustice to the Dow Theory but just the contrary: I am strictly qualifying trends and secondary reactions in real time as per the rules of the Dow Theory.

The duration of primary bull and bear markets is not carved in stone. We have empirical guidelines, but we should react in real time to the signals the market is displaying. Furthermore, as I will show below, there are many instances of primary bull markets that lasted less than a year, and more, importantly, investment/trades following the Dow Theory entry and exit signals that lasted significantly less than a year.


You have taken considerable care to justify your frequent calls of a primary bull and primary bear market with the following comments on January 2, 2013 (found here:

“I am very aware that recently this Dow Theory ‘flavor’ has signaled too many signals in too short time. On June 4, 2012 (even though I was not blogging at that time), the Dow Theory flashed a primary bear market signal which was negated by a subsequent primary bull market signal on June 29. Later, on Nov 16, a primary bear market was signaled (11/07/2012 as per Schannep's). Some critics would say that such short-lived signals do not bide well with the claim that primary trends tend to last 1-2 years on average.”

However, it would be hard to claim that “critics” are not in favor of such short duration primary moves in the market when those same individuals are deriving their view from the primary & secondary sources. Especially when the “criticism” is based on direct quotes from Charles Dow himself. The following are Dow’s own words on the concept of a primary market move:

“We have frequently demonstrated that the stock market, while full of short fluctuations [also known as secondary reactions], has a continuing main movement [primary bull/bear market], which often runs in one direction for three or four years at a time.(source: Dow, Charles H. Review and Outlook. Wall Street Journal. September 13, 1900).”


While understanding Charles Dow is a must for any Dow Theorist, subsequent Dow Theorists have made clear that Charles Dow expected too long primary and bear markets. We should bear in mind that Charles Dow had a much shorter empirical record than us. This is why Rhea wrote that the Dow Theory would gain in usefulness as time passes. The longer the track record, the more observations (empirical record) we have. Thus, primary bull markets, even under the “classic” Dow Theory do not necessarily have to last 4 years on average.

Furthermore, the “classical” Edwards and Magee Dow Theory record supports the view that investments made along primary bull market signals do not last on average 4 years. Thus, the Dow Theory record from 1897 to 1956, registers 15 buys and 15 sells, which results in one transaction every 2 years on average, which is clearly less than 3-4 years. I insist this is with the “classical” Dow Theory.


To proffer Jack Schannep or E. George Schaefer’s theory for your method of interpreting the markets is great, especially when it works.


I feel I have been misread. I am 100% follower of Schannep’s. However, I am in the antipodes of Schaefer as he is an investor along the secular trend and I always make clear that I am not interested in the secular trend. Future posts will expose the dangers inherent with blindly following the secular trend and ignoring the warnings of the somewhat shorter term primary bull and bear markets as defined per the Dow Theory.

I agree that blindly following the secular trend as Schaefer suggests can be lethal to the investor. This is why; in the first place, I am not a follower of Schaefer at all and the more I think I know about markets, the more I see the hidden danger in riding without stops secular trends. I find value in Schaefer, though, when he helps us determine the secular trend. However, I use the secular trend only in order to know which asset class is more likely to face tail or headwind and, accordingly, allocate capital differently. I agree that Schafer’s approach looks great…when it works.

However, we cannot mix Schaefer with Schannep. By definition, Schannep’s version is very reactive to market action, as it is "classic" Dow Theory  (which in itself is very responsive) with improved responsiveness. Thus, it is highly unlikely that the investor will be decimated by overstaying a bear market (unlike those investing along the secular trend) either with the "classic" Dow Theory or with Schannep's version. 

Finally, there is no known method of investment that is always right. The holy grail doesn't exist as any seasoned investor knows. Sometimes,  the classical Dow Theory will outperform Schannep and sometimes  not. However, over long periods of time (i.e. 5 years) Schannep’s tends to outperform the classical Dow Theory. More importantly it seems that Schannep’s flavor results in shorter and less pronounced draw-downs. I insist no system or “flavor” can be right 100% of the time. The goal is to scratch some out performance while, more importantly, keeping drawdowns as small as possible. The classical Dow Theory does an excellent job at this and Schannep, to the best of my knowledge, even improves something that was close to being perfect.  

Readers of this Dow Theory blog, stay tuned, because in the coming days I will post the result of some in depth research that clearly shows that even the "classical" Dow Theory under performs most of the years buy and hold. So Schannep's is not to blame. However, the conclusions of the study are trully astounding and confirm that the Dow Theory, even the "classical" one, does a brilliant job in keeping investors safe. While under performing buy and hold quite often, the Dow Theory manages to have smaller draw downs and greater performance because out performance comes when most needed: When the market is falling big time. So the Dow Theory acts as a king of "insurance" against bear markets.


However, to suggest their strategy is Dow’s theory (as modified and then codified by Robert Rhea based on the works of William Peter Hamilton) brings legitimacy to the criticisms against Dow Theory.


Schaefer and Rhea are in the antipodes. So it escapes me lumping them together. Precisely, Schaefer in a not so subtle way overtly criticized Rhea, as you can read  here.

Of course, if the criticism is intended at discrediting Rhea and say that the only pure Dow Theorist was Charles Dow and that Hamilton and Rhea (and by implication, Russell and Schannep) “spoiled” the original Dow Theory as expounded by Charles Dow this is an open question, and NLO is not the first one to make such a claim. Schaefer in his book openly criticizes Rhea and Hamilton for having turned into a mechanical system the wisdom contained in Charles Dow writings.

However, I take such criticisms with a grain of salt, since it is beyond doubt that Rhea was an excellent investor. Here maybe I am too practical but if something works why argue? Even accepting Schaefer’s criticism that Rhea degraded the Dow Theory by turning it into a set of rules and making it shorter term oriented (as if 1-3 years were too “short term oriented), it is beyond doubt that Rhea’s Dow Theory flavor works. The same applies to Schannep, who, to the “orthodoxy’s” dismay, has a proven track record against which it is difficult to argue.

Thus, I think it is a bit of overstretch to proclaim that only the writings of Charles Dow deserve the name of “Dow Theory” particularly when Dow himself never attempted to label his writings as “Dow Theory”.

As I wrote here:

Is any one of the Dow Theory “flavors” we have hitherto analyzed to be excommunicated from the Dow Theory fraternity? Personally, I think that all of them loosely fit under what we can define the basic Creed of the Dow Theory. Some styles place greater emphasis on the secular trend and values; others focus on riding the primary trends and are more technical in nature.


Worst still, RETROCASTING, referring to old data and changing your current view to fit what has already passed in contradiction to the real-time call you’ve made, is misleading at best.


As I explained above, ad nauseam,  such statement is not correct. Nothing has been retrocasted.


Another concern is the rationalizing of the idea that “…’average’ means ‘average.’ Thus, it is not expected that every Dow Theory primary market is going to match exactly the alleged ‘average’. Some primary markets will last longer than average (i.e. 1-2 years in bull markets; 6 months in bear markets) whereas others will last significantly less. So it is possible to see bull markets exceeding 6 years as it is possible to see bull markets fading after some few months.”

Referring to the “bible” of technical analysis, Technical Analysis of Stock Trends by Edwards and Magee (1991 edition. Page 61 found here:, Dow Theory NEVER had a primary bull market that lasted less than a year from 1897 to 1957. According to the book Trader Vic II by Victor Sperandeo, Dow Theory had only two primary bull markets that lasted less than a year (3/31/38-11/12/38 and 4/8/39-9/12/39) from 1896 to 1989 (1997. Page 106). According to Jack Schannep’s book Dow Theory for the 21st Century, it is stated that the period from 1897-1956 had only one period when a primary bull market was less than a year (7/17/1939-5/13/40) [This data was derived from the 2001 edition of Technical Analysis of Stock Trends].


Following clarifications are in order:

1.    NLO is mixing total duration of a bull market with total duration of the investments made following primary bull market signals. While similar it is not the same and, by definition, one buys after a primary bull market signal has been signaled (which always happens with some delay and not at the absolute market bottom). By the same token, one exits the trade when a primary bear market has been signaled, which always happens with some delay and not at the bull market’s top. Thus, depending on each specific instance, the time invested in the market will be shorter or longer than the primary bull market itself.

2.    Thus, Sperandeo’s classification of bull markets is providing us with the total duration of a bull market, not with the specific entry and exit price and date of the investments made following the Dow Theory. Furthermore, and while not disagreeing with Sperandeo, his classification lacks the clarity provided by Schannep in his book “Dow Theory for the 21 First Century” (pages 54-58) whose classification of bull markets provides us with a wealth of details as to their their duration and extent.

3.   It is not correct to say that the period from 1897-1956 had only one period when a primary bull market was less than a year (7/17/1939-5/13/40). I suppose NLO really means “there was never a trade according to the Dow Theory that lasted less than a year." As I have written above, we should not confuse total bull market duration with trade duration. I agree with NLO as to the source (Schannep and Technical Analysis of Stock Trends). However, if we go to page 29 of Schannep’s book (Table 3.1, 2008 Edition), we can see that there was another instance that lasted less than one year, namely 6/23/1938-3/31/1939. Of course, shorter trades are closely aligned with shorter bull markets. 


However, under Schannep’s personal analysis of Dow Theory primary bull markets from 1954 to 1999, there were six periods when Dow Theory primary bull markets occurred in less than a year. The dates and performance are as follows:

10/10/1961 to 4/26/1962: -3.96%
1/11/1967 to 10/24/1967: +7.99%
10/1/1968 to 2/25/1969: -4.51%
9/28/70 to 7/28/1971: +14.89%
6/6/1978 to 10/19/1978: -2.32%
6/3/1990 to 8/3/1990: -4.28%

Not surprisingly, the “average” performance of those periods that were less than a year was +1.30%. This suggests that the above dates need to be studied to understand how they were not accurate for the purpose of avoiding such short term indications.


It is true that from 1956, the “classical” Dow Theory track record was reconstructed by Schannep. However, Schannep did not reconstruct such record with his own Dow Theory “flavor” but he limited himself to applying the classical Dow Theory rules.

Furthermore, as Schannep says, “the Dow Theory is not an objective algorithm” and thus, even under the classical Dow Theory, it is impossible to say that there is only one valid interpretation. Thus, I find strongly biased against Schannep to dismiss his reconstruction of the “classical” Dow Theory record, because I find that his analysis was not “personal” but faithfully reflects the classical Dow Theory.

Furthermore, I am in possession of Russell’s “Dow Theory Letters Technical Studies” which contain the detailed price action of the Industrials, Rails (Transports) and Utilities since 1885 to date. I have personally checked on the charts the “classical” Dow Theory signals and, to the best of my knowledge, Schannep has made a rigorous application of the “classical” Dow Theory.

By the way, such short-duration trades confirm my view that during secular bear markets, primary bull markets have a shorter life span. It is not by chance, that such spell of abortive primary bull markets occurred during the secular bear market that spanned from 1966 to 1980. Only two “short-lived” primary bull market occurred outside the secular bear market. 

Even if Schannep's interpretation of the "classical" Dow Theory was flawed (which is not, to the best of my knowledge), we must compare apples to apples. It is disingenuous to claim that in the midst of a secular bear market, Schannep's interpretation of the Dow Theory "only" managed to achieve +1.30% when the overall stock market experienced draw downs that reached 50% during such secular market period. Furthermore, since this +1.30% was achieved in less than one year, annualized it is certainly more, something in the vicinity of +2% (plus the interest earned during the periods one was in cash). So bearing in mind the existence of a secular bear market, the "wrong" interpretation of the classical Dow Theory was not that bad after all. I wish I could secure ca. 2% performance when the next secular market decline of 50% hits. However, I would gladly be, together with Schannep, proven wrong and be shown that the "right" Dow Theory interpretation resulted in extracting in the midst of a secular bear market stellar returns. I don't have pride of opinion, but pride of increasing the bottom line.

Furthermore, the real test of whether Schannep misapplied the classical Dow Theory, and, thus, in the 1961 to 1978 period (when 5 out of 6 "less-than-one-year" investments occurred), Schannep's "misapplication" resulted in sub par performance is to take the whole period under consideration. The figures I will give right now, while deemed accurate are still in process of verification, but it is worth anticipating some results of our ongoing research.

During the 1961-1978 period buy and hold returned + 2.93%. Schannep's reading of the classical Dow Theory returned +4.53%. However, this is only part of the story. During this period, the standard deviation of returns for buy and hope, sorry, hold was 17.44% whereas Schannep's interpretation of the classical Dow Theory was a modest 11.11%. Furthermore, in killing years such as 1973 and 1974, holding the Industrials would have resulted in monstrous losses whereas the classical Dow Theory, as interpreted by Schannep, didn't even reach losses of -6%. (all these figures are on a year end basis).

Therefore, with all due respect, I believe Schannep's criticism is unfounded. He certainly did a good job in reconstructing the "classical" Dow Theory record from 1956 to date.

All in all, neither Schannep, nor Rhea or yours truly should be discredited because his Dow Theory “flavor” resulted in primary bull and bear market signals that lasted less than one year. This has happened before (even with the “basic” Dow Theory flavor), happens now and is likely to happen when cyclical bull markets occur in the midst of a secular bear market.


The review of the 3 different sources also tells us that Schannep’s approach is probably the least robust method to apply. After all, under Schannep’s approach, 5 of 18 Dow Theory primary bull markets generated negative returns by the time a “sell” signal was given. This means that only 72% of the bull market indications were correct. Contrast that with Sperandeo and Edwards & Magee’s 100% of bull market indications generating positive returns.


I feel that NLO is mixing its dislike for Schannep’s own flavor with Schannep’s faithful reconstruction of the Dow Theory record as per the “classical” rules. 

Furthermore, we have to compare apples to apples. As I have written above, Sperandeo bull market tabulation merely reflects the total gain experienced by bull markets (hopefully as defined by the Dow Theory which, by the way, is not so clear to me, as he doesn't provide the wealth of details given by Schannep in the Appendix A of his book). Of course, if we tabulate "bull markets" by definition all of them are positive. However, it is not the same to say that a bull market gained 40% from bottom to top than "extracting" gains therefrom. By definition the buy signal will be after the bottom has been made and the sell signal after the top has been reached. Thus, it is perfectly possible to have bull markets that nevertheless result in losses to the investor. Thus, it is wrong to compare Spearandeo's with Schannep's record.

As to the Dow Theory record until 1956 (the one made by Edwards & Magee) which is more like an apples to apples comparison since they deal with entries and exits ("net" performance versus "raw" performance with Sperandeo) it is wrong to say that they had a 100% batting average. In 1939-1940 there was a trade that resulted in a loss of -3.6% and had their Dow Theory record being extended until 2012, it would certainly have suffered occasional losses.

I can understand that Dow Theorists of the classical or even pure Charles Dow persuasion may not like Schannep’s approach (i.e. including 3 indices). However, I feel that this dislike for Schannep’s innovations results in discrediting anything Schannep has to say even when Schannep merely is applying the “classical” Dow Theory.

I insist: 4 of these short lived bull markets occurred in the midst of a secular bear market. I have checked on the charts the Dow Theory signals and while this is not math, I can only say that I feel that Schannep has been rigorous in applying the classical Dow Theory.


Now, Schannep seems to believe that his interpretation is rigorous enough to withstand the test of time while generating positive returns to the investor. However, even if the Sperandeo and Edwards and Magee interpretation of Dow Theory are simply ex-post analysis, or RETROCASTS, wouldn’t it be better to see if their analysis is correct since we know for certain that Schannep’s 72% is inferior to the alternatives mention earlier.


Once again, I have to point out that NLO is mixing total duration of a bull market with total duration of the investments made following primary bull market signals.  Sperandeo’s figures (Table 9.2 of his book) refer to bull markets from bottom to top. So the percentage gain showed do not correspond with the gains that an investor would make according to the Dow Theory. In any primary bull market swing ca. 20% thereof is lost to the investor because the buy signal tends to occur ca. 10% after the bottom was made and the sell signal comes ca. 10% after the top has made. Thus, Sperandeo is giving us “raw” measures of performance.

Furthermore, and while not discrediting Sperandeo because this is not math, there are discrepancies between the bull markets recorded by Sperandeo and Edwards and Magee. Thus, Sperandeo writes that 1947-1948 was a bull market that lasted 409 days with a gain of 38.1% whereas Edwards and Magee (Table 3.1 Schannep’s book) show that during this period the investor was flat suggesting no bull market had been detected by the Dow Theory.

Furthermore, I'd be extremely leery of a technical system that promises a 100% batting average. Then, everybody would be following it, and it would self destruct. Nonetheless, as I said, such claim is wrong. Sperandeo give us "raw" performance, whereas Schannep's record give us "net" performance.

Moreover, Sperandeo didn’t write a book about Dow Theory but about investment in general. The Dow Theory is merely another tool in his arsenal. There is nothing wrong with this, but the astute investor should notice the intense focus or a Rhea, Russell or Schannep (devoted Dow Theorists) with other investors who, legitimately, write about the Dow Theory in a much less intense fashion.

So I feel it is a bit of a stretch to canonize others while excommunicating Schannep even when Schannep is merely trying to apply the “classic” version of the Dow Theory.

The good thing in investing is that it is quite easy to show who is right and how is not so right. In investing right are wrong are measured by the bottom line. Here you can find Schannep's track record so that you can judge for yourself.

And when it comes to the bottom line, Schannep shines.


Again, this is simply an examination of Dow Theory as students of the stock market and the relevant sources on the topic. Your discussion of Dow Theory is certainly stimulating and I hope that the above points only further the discussion.


Thank you and I also hope that my clarifications are taken in this spirit so that we all advance in our knowledge of the Dow Theory and become more successful when investing.


No comments:

Post a Comment