Dow Theory and the Gold Stock Indicator

Reader FH asks:

“It appears that on December 2, 2013 the Gold Stock Indicator may have fully retraced its level from the end of June? [I] believe you had mentioned that this was predicted by Dow theory and was needed for the indicator to bottom? Can you elaborate on the implications of this move in the context of Dow theory?”

Our Response:

In order to accurately answer this question, we must review Dow Theory on the concept of retesting prior lows.  To do this, we’re going to enlist the more insightful writing of Richard Russell at his very best on the topic of Dow Theory (note: the idea of retesting/retracing is the equivalent to Dow Theory’s concept of double bottoms and double tops). First, Russell addresses the matter of what is essential for the reversal of a bear market trend in the following commentary:

"PRIMARY CHANGES As this Letter is written, Industrials are 16 points below their peak at J, Rails 4.5 points below their comparable J high. Bullish views are usually discussed to best advantage (sans emotions) in a declining market. For this reason, I will use the following section for a survey of the bullish possibilities.

"Robert Rhea recognized two formation as signifying the turn from a bear to a bull market. These two formations are shown on the accompanying chart. In Figure 1a both Averages decline to bear market lows at point A. Following A, a rally commences (A-B) which, for want of more conclusive evidence, we classify as an upward reaction in a continuing bear maker. A decline (B-C) ensues which lasts at least two or three weeks and which corrects about 1/3 to 2/3 of the A-B advance. On the B-C decline one or both Averages hold above the previous lows at A. It is what occurs after C that is important. If, on the next advance. both Averages better the preceding high (at B) the primary trend is taken to have turned from bear to bull. In this case we go back and reclassify the A-B advance as the first primary leg in a new bull marker; the B-C decline then becomes the first reaction in the new bull market, and the advance from C becomes the second primary bull leg.

"In figure 1b we see a more unusual formation and one which is in many ways less satisfactory. Nevertheless, Rhea accepted the formation in Figure 1b as implying a change in the primary  direction. On the decline to A, ‘bear market lows are recorded. An upward reaction ensues (A -B), followed by a further decline which takes both Averages to new bear market lows at C. At C there is little to suggest that the primary trend has turned (except, perhaps, for a drastic diminution in volume). Subsequent to the C lows both Averages turn up and advance steadily (with little or no correction). At the point where the preceding secondary high (B) are bettered by both Averages the primary trend is considered to have changed from bear to bull." (source: Russell, Richard. Dow Theory Letters. February 13, 1961. Issue 133. page 2.)

The literal plotting of the way a primary trend reversal should look clarified for us many of the questions we’ve had about what to expect or how to anticipate a potential reversal of the trend.  However, there are some key points that Russell makes in the discussion above.

1. Bullish views are usually discussed to best advantage (sans emotions) in a declining market.
2. Robert Rhea recognized two formation as signifying the turn from a bear to a bull market.
3. Rhea accepted the formation in Figure 1b as implying a change in the primary  direction.
4. At the point where the preceding secondary high (B) are bettered by both Averages the primary trend is considered to have changed from bear to bull.

Regarding #1, a reasonable and balanced expectation of a trend reversal within a declining trend cannot be accomplished with excessive emotion. Regarding #2, there are two types of “double bottoms” that can occur that would indicate that the trend may be about to reverse.  However, in both examples (Figure 1a and Figure 1b) a retest of the previous low should be expected. Regarding #3, even though the market goes materially below the prior low does not necessarily mean that a continuation of the bear market has been signaled.  The success or failure to exceed the prior secondary reaction at point B in Figure 1 is the best indication of new bull or bear market as suggested in #4.

To better understand this concept we must point out what should be the glaring example of when the above scenarios don’t work.  In our Gold Stock Indicator (GSI), you can see the period from May 2012 to September 2012 where the GSI did exactly what is indicated above but ultimately failed (indicated by the red A, B and C).

What this tells us is to be careful about over commitment to such a theory as it might not work out.  Our treatment of this failure in the theory was to commit small amounts initially and increase our commitments over time and with additional confirmation.  Even in our confidence of the theory, we gave a sell recommendation of Agnico-Eagle Mine (AEM) after a +45% rise from our April 2012 recommendation which was just short of the November 2012 (found here) peak; AEM now sits –48% below the September 2012 sell recommendation.  This recommendation to sell was in keeping with Dow’s view of “seeking fair profits (found here)”  The theory is only as good as our ability to acknowledge and accept its weaknesses and our willingness to work around such deficiencies.

Our next example from Richard Russell’s Dow Theory Letters has the following thoughts:

"COMPARISONS. A study of the history of the price movement suggests that the more times an important point has been 'tested' (approached but not violated), the more drastic the resultant move if and when that point is penetrated. The accompanying chart traces the pattern of the Averages throughout the year 1907. Because of the similarities in the price movement then and now, I feel that the 1907 period is well worth studying. Following a sharp decline which began in late 1906, both Averages touched lows in March, 1907 (A on chart). From there a rally extended into early May (B). In mid-May the two Averages declined to test the previous March lows (C), then rallied to point D. From D another decline (to E) served to retest the earlier C levels. When the C levels again held, the advance to F was sparked.

"Observers in July, 1907, must have considered the A-C-E area almost impenetrable. But a great bear marker was in progress, and the primary trend continued down. In October the 'triple base' was shattered, and the force of the penetration was breath-taking. In a period of a few months Industrials lost 30% of their value.

"The 1907 example was not unique. Only five years earlier, in 1902, a double bottom was broken, following which one of the most drastic declines in record set in. Because of these and other precedents, it seems to me that if the current bear market lows are violated probabilities again favor some drastic downside action." (source: Russell, Richard. Dow Theory Letters. July 22, 1960. Issue 111. page 3.)

The demonstration of failures teach more than the actual successes.  Failures teach us to remain ever vigilant in the use of such theories while at the same time taking advantage of the many times when the theory works.  In the example above we learn a few things about the importance of retesting prior lows:

1. the more times an important point has been 'tested' (approached but not violated), the more drastic the resultant move if and when that point is penetrated.
2. Point B needs to be violated on the upside to consider the primary trend bullish.  The failure of point F to exceed point B was an key indication that the trend was still down.
3. A “Line” of accumulation/distribution is established when price trades in a range; violation above or below that range determines the primary trend.

Regarding #1, testing a prior level, either at the high or low, over an extended period of time indicates that the subsequent move may be very powerful. Regarding #2, not being able to exceed point B was the clearest indication that the a retest of A-C-E was in order.  However, there was no proof that the subsequent substantial decline below A-C-E was inevitable.  As observers and investors of financial markets, we seek confirmation but should have accounted for the prospect that we would be wrong well in advance.  Regarding #3, our best summation of the concept of lines is found in the following quote:

“Since November 9, 2009, the Dow Industrials have traded in a tight range of less than 3%. According to Dow theory, the market trading in a range of about 5% is considered to be a “line.” A line is a key indication of accumulation or distribution of stocks. It is not known whether or not accumulation or distribution has taken place until the market either breaks above the high range or below the low range of the line.

“According to Dow Theory, the formation of a line can act as the equivalent of a market decline or secondary reaction in a bull market if it lasts for over eight weeks. In this instance, the line lasted exactly 8 weeks. I’m hesitant to accept that the “rule” of 8 weeks can be trusted altogether. However, the upward bias of the market has indicated that the most recent breakout will be followed until proven otherwise. It is important to note that secondary reactions act as a release valve from built-up pressure in the market. The fact that the market has responded by breaking above the line that had been drawn indicates that the market has successfully absorbed the large amount of shares that have been distributed by corporations as well as the negative economic news since the March 2009 low.

“Because the Dow Industrials and the Dow Transports have both broken to brand new highs at the same time, along with the fact that the Dow Industrials have broken above the line that has been drawn since November 9, 2009, we can safely guess that the market has little desire to go lower and that the bull market is still in place. (found here)”

There are two more comments by Russell that we should take into consideration on the topic of retesting prior low and their implications:

"...before you can get a bull move, you have to establish a strong base -- and you have to test and retest that base many times." (source: Russell, Richard. Dow Theory Letters. October 18, 1989. Issue 1028. page 6.)

"...Gold popped up today, and I’m getting calls from all over the world with the question, 'Do I buy here?' I would prefer to hold off because after yesterday’s panic drop below 130, it would be normal to get a bounce halfway back to 160 or a move say to 145 or thereabouts. But just as surely after the first bounce off a panic low, we should see a retesting of the lows and usually a break below the first panic low. Buy a little gold here if you want (no shares), but expect a retest of 126 with very likely a drop to the ll5-120 area within a few months." (source: Russell, Richard. Dow Theory Letters. September 26, 1975. Issue 643. page 7.)

In summary, retesting the prior low and possibly going lower, is necessary before a bull market run can be firmly established.  Indications after the retest will be key to the future direction of the GSI.  Of additional concern is the fact that when the general equity market declines so too does gold and gold stocks by a greater magnitude as indicated in many articles in the past.  If the critics of the market are right that we’re in a bubble and said bubble does burst then gold stocks will get decimated beyond recognition.

There are many flaws in the process when interpreting double bottoms from a Dow Theory and real world investment perspective.  However, the concept of double bottoms and double tops also known as retesting/retracing in the Dow Jones Industrial Average or Dow Jones Transportation Average have accounted for 72% of the major bull and bear market turns in the stock market. Focusing only on the successes will render any knowledge on the topic meaningless and foster an unhealthy perception of risk.  Again, the theory is only as good as our ability to acknowledge and accept its weaknesses and our willingness to work around such deficiencies.