Dow Theory Deception
Like a raging inferno engulfing a dry prairie, a rumor has been sweeping the online financial community of late.
The rumor concerns the recent low made by the Dow Jones Transportation Average and the supposedly bearish implication it has for the stock market in the coming weeks and months.
It is being hailed by many as a "Dow theory sell signal" and is being talked about across many financial publications and Internet sites as the latest in a series of bad news for the broad market. The thinking behind this heavily promoted fear is that the Dow Jones Transportation Average made a lower low in November against its August correction low. Therefore a bear market is said to be imminent.
Does this qualify as a legitimate Dow theory sell signal? Is a bear market and economic recession coming our way? Before we can find the answers to these questions, it might help to refresh our memories as to the meaning behind Dow theory.
The Dow theory was developed by Charles Dow, founder and editor of the Wall Street Journal, in 1897. Dow developed the two well-known broad market averages, the Rails index (later the Transports) and the Industrials. The original rail average was comprised of 20 railroad stocks.
The Dow Theory resulted from a series of articles published by Charles Dow in the Wall Street Journal between 1900 and 1902 and later popularized by his successor, William Peter Hamilton. Hamilton published a book on Dow theory entitled "The Stock Market Barometer." Later authors such as Robert Rhea gave further prominence to Dow's theory.
The basic tenets of the Dow theory are as follows:
- The two major averages (industrials and transportation) discount everything.
- The market has three trends: primary, secondary and minor.
- The averages must confirm each other.
- A trend is assumed to be in effect until it gives definite signals that it has reversed.
With these rules in mind, let's have a look at what some of the pioneer Dow theorists had to say about the theory and its interpretation.
First, it will do us well to remember that Robert Rhea, in his classic book "The Dow Theory," gave the following disclaimer: "The Dow theory is not an infallible system for beating the market. Its successful use as an aid in speculation requires serious study, and the summing up of evidence must be impartial. The wish must never be allowed to father the thought."
Referencing one of Hamilton's Wall Street Journal editorials, he writes, "The market does not trade upon what everybody knows, but upon what those with the best information can foresee. There is an explanation for every stock movement somewhere in the future, and the much talked of manipulation is a trifling factor." (Jan. 20, 1913)
Hamilton further wrote, "It is much harder to call the turn at the top than at the bottom....The market, moreover, perhaps because of the complexity of the situation and more truly because of the stability of the general prosperity predicted by the barometer, may hold within a relatively small distance from the top for an indefinite time. It might indeed be said that there are instances of nearly a year with a range not far from the top, before an aggressive bear market has been established." (Feb. 15, 1926)
These are important nuances to remember when approaching the Dow theory. A sell signal is a major event that isn't to be treated lightly and Dow himself would almost certainly disapprove at the way his theory has been misapplied in making bear market predictions.
In further disclaiming the theory's infallibility, Hamilton said, "The task of calling the exact top to a major movement is beyond the scope of any barometer. It is additionally difficult where there has been no inflated speculation....Indeed, it may almost be said that a bear argument understood is a bear argument discounted."
In "The Dow Theory," Rhea observes, "The movement of both the railroad and industrial stock averages should always be considered together. The movement of one price average must be confirmed by the other before reliable inferences may be drawn. Conclusions based upon the movement of one average, unconfirmed by the other, are almost certain to prove misleading."
This comment can be applied to the current discussion of the supposed Dow theory sell signal. As we've already read in the above, the Dow Transportation Average gives its best signals when it confirms the Industrials. Just because one index makes a lower low while the other doesn't should not be interpreted as a sell signal at face value. History shows many instances where the Transports made a lower low at a correction bottom, only for the Dow Industrials to go on to make higher highs.
The October 1998 correction low was one such instance. Back then, the Industrials made a double bottom while the Transports made a lower low. If a trader was looking at the Transports for a directional clue in October '98, he would have been whipsawed as the broad market went on to recover its losses into December '98.
Another instance of the Transports making a lower low against the Industrials was in 1994. The DJTA actually made a series of lower lows between May and December of '94 while the DJIA made slightly higher lows during that same period.
If an investor took his cue from the Transports (as many unfortunately did), he'd have missed one of the most vibrant and sustained broad market rallies of all time in 1995.
The broad market doesn't necessarily need the participation of the Transports in order to experience a bull market. A recent example of lagging performance on the part of the Transportation stocks would be the 2006 experience.
Last year, the Transports peaked in May and didn't recover its previous high until February 2007. The Dow also peaked in May '06 but recovered its high in September '06 and went on to make higher highs through February '07. That's a lag of several months before the Transports eventually caught up with the Industrials.
The granddaddy of Dow Theory non-conformation signals would have to be the incident of the late 1880s through the early 1890s. Why bother to bring this up since it was so long ago? Because it's precisely analogous to our own time. You see, back then the 120-year Kress Cycle was bottoming into the mid-1890s. We are heading into the Kress Cycle bottom of roughly 2014. That means the period of 2007-2014 closely corresponds with the period of 1887-1894.
Let's look at what happened in 1887-1889. The equivalent of the Transportation Index, namely the 20 Rails Index, peaked late in 1882 and declined sharply until 1885. From there it fluctuated in a relatively narrow, lateral range until the early 1900s. From 1887-1889 (equivalent to 2007-2009) the Rails went sideways. (Source: A Century of Prices, by Sen. Theodore Burton and G.C. Selden, 1919).
Compare this to the equivalent of the Dow Jones Industrial Average of 1887-1889. The Axe-Houghton Industrial Stock Price Average took a hit in mid-1887 but bottomed out by the fourth quarter and consolidated into early 1888. Then it took off and had a stellar second half of '88 and continued rising into 1889, peaking out at the top of a long-term uptrend channel in early 1890. The uptrend in the Industrials didn't actually end until early 1893.
This shows there can be a multi-year lag between the Transports and the Industrials in extreme cases.
Rhea himself called the confirmation of the two averages "the most useful part of the Dow theory, and the part that must never be forgotten for even a day."
Hamilton wrote, "Dow always ignored a movement of one average which was not confirmed by the other, and experience since his death has shown the wisdom of that method of checking the reading of the averages." (June 25, 1928)
What is the inference of a stock market movement where the averages don't confirm each other? According to Hamilton, "Uncertainty still continues as concerns the business outlook..." (May 24, 1924) It doesn't guarantee a bear market or recession is around the corner, nor does it necessarily constitute a bona fide sell signal. It just means the business outlook is "uncertain" in pure Dow theory terms.
Hamilton went on to state, "There is one fairly safe rule about reading the averages, even if it is a negative one. That is that half an indication is not necessarily better than no indication at all. The two averages must confirm each other.." (Aug. 27, 1928)
He states further, "Indeed it may be said that a new high or a new low by one of the averages unconfirmed by the other has been invariably deceptive." (May 10, 1921)
Here are some further observations of Hamilton as recorded by Rhea:
"...when one breaks through an old low level without the other, or when one establishes a new high for the short swing, unsupported, the inference is almost invariably deceptive." (Feb. 10, 1915)
"...conclusions drawn from one average but not confirmed by the other, are sometimes misleading, and should always be treated with caution..." (June 26, 1925)
"Once more it is worth while to emphasize that the movement of these two averages is deceptive unless they act together." (June 9, 1915)
"They are frequently misleading where one group breaks through the line [of support or resistance] and the other does not. When, however, the movement is simultaneous there is a uniform body of experience to indicate the market trend." (April 16, 1914)
"One of the shortest ways of going wrong is to accept the indication by one average which has not been clearly confirmed by the other." (The Stock Market Barometer)
There you have it, wisdom straight from the mouths of the Dow theory's founding fathers.
The Dow theory as conceived by Dow and Hamilton was never intended to be a standalone trading tool. It was simply a method for understanding overall business conditions in the U.S. Modern day practitioners are sometimes inclined to make entirely too much of Dow's theory in the way of stock trend predictions.
Dow theory is too cumbersome a tool to be used as a reliable trading system, its signals often misleading. Indeed, those that have tried to do so have found it wanting. To be useful to an investor it must be supplemented with other forms of analysis, both technical and fundamental.