Confirmation Analysis

As the Dow Jones Transportation Average (DJTA) pulls back from all-time highs in February 2005 and the Dow Jones Industrial Average (DJIA) struggles to keep up, Dow theorists are at a juncture no less important than the one they found themselves back in 2000 when the transports topped out in April, only to be followed by the industrials some eight months later. Here in 2005, the question of the Dow averages and all-time highs once again has cropped up, and can be summarized as this: The transports made a new all-time high in November 2004, while the industrials have yet to confirm with a new, all-time high of their own. If the industrials do make a new, all-time high (some 740-odd points as of this writing), then will the ballgame be completely over for the secular bear market thesis?

A technical analyst and student of the market's cyclical behavior, Tim Wood of Cyclesman.com has done technical analysts worldwide a great service by reminding them that while intermediate-term buy and sell signals may have occurred since the 2000 "master sell" signal, it began the bear market. Wood explains the master sell signal this way:

As the Industrials moved into July and August of 1999, they were unconfirmed by the [t]ransports... Then, in September 1999, both averages moved below their previous secondary reaction low points... This in turn produced a Dow theory sell signal.

However, from the October 1999 low, the Industrials were able to make another new high as they advanced into January 2000. This new high was also unconfirmed by the [t]ransports. From this unconfirmed high, both averages broke once again below their previous secondary low points... This served to reconfirm the 1999 sell signal as well as to mark the beginning of the first leg down of Phase I of this great bear market.

And in case anyone is still unclear on the point:

The 1999 sell signal is a Primary Dow theory sell signal. In simplest terms this means that any price action, occurring after the primary sell signal, is subordinate to that signal, until/unless the Primary signal is corrected.

It is on the precipice of just such a correction that the markets — and those trading them — find themselves in the first few months of 2005. To this end, it seems as if there are three questions that seem paramount in trying to understand the relationship between the transports, the industrials, and the likelihood of a rising or falling market going forward in 2005:

1 Did transport weakness anticipate the secular top in 2000?

2 Did transport strength anticipate the cyclical bottom in 2002?

3 Is lagging transport performance in 2005 suggestive of a cyclical top?

Three not-so-easy pieces

Here I'm veering away from textbook Dow theory analysis of confirmation between the transports and industrials to look instead at each average on its own technical merits. With a master sell signal on one end (1999-2000) and the potential for a master buy signal on the other (2005), I will use basic technicals to both support the (now-obvious) case for the master sell signal in 1999, and to highlight those technical factors that may or may not contribute to a master buy signal in 2005.

Let's look at the first of these three pieces. Technically speaking (and, again, I'm avoiding specific Dow theory interpretations), the transports developed a negative stochastic divergence on a monthly scale back in 1997­98. However, it appears as if this monthly negative stochastic divergence was anticipating not so much The Top as the market break that was the minicrash of 1998 surrounding the Russian debt default, the Long Term Capital Management financial implosion, and lingering problems from the Asian financial crisis.

The transports rallied from the lows of 1998 to make a new, all-time high in April 1999. However — and again focusing on monthly charts — we can see that the transports did not show followthrough to the upside in May, June, or July, when the transports in fact reversed themselves in earnest. This kind of price action is identical to that highlighted by traders as the 2B top (Victor Sperandeo) or Turtle Soup (Larry Connors) setup.

This setup, about which I have written frequently in Working-Money.com and Traders.com Advantage, suggests that when a market makes a high, then a higher high, yet fails to show followthrough to the upside in the wake of that higher high, then a significant chance exists that the market is ripe for reversal. And by the middle of 1999, a reversal is exactly what the transports had in store.

Figure 1: An intermediate 2B top in the late 1990s led to the bear market of 2000-02. At present, a long-term potential 2B top between 1999 and 2005 looms over the transports.

Note that this sell signal/reversal signal in the transports arrived several weeks before the September Dow theory sell signal, months before that master sell signal was confirmed in March 2000. It is also worth mentioning that while the transports produced a 2B top/Turtle Soup sell signal in the summer of 1999, there was no way of knowing how significant the sell signal was.

Save for the fact that it appeared on a monthly chart, and thus was certainly not a short-term signal, the 2B top/Turtle Soup­based sell signal did not provide the same sort of "long-term" assurance that the Dow theory-derived sell signal and confirmation in September 1999 and March 2000 did. In the absence of Dow theory, the next buy signal in the monthly chart of the transports would potentially be considered as significant, meaningful, and tradable as the sell signal in 1999.

How useful would this individual analysis of the transports have been to a trader in 1999? Because the strict Dow theorist requires confirmation in the behavior of both averages, a sense of how likely a given move in one average might be could help a trader anticipate the likelihood of a confirmation. A trader sensitive to the bearishness of the transports in mid-1999 would have been skeptical of the Dow's continued advance that year — all the while unloading or shorting transportation stocks.

Transports and cyclical turns

So if weakness in the transports anticipated the secular turning point in 1999-2000, did DJTA strength anticipate the cyclical bottom in 2002? As you might imagine, my answer here is "yes."

Sometimes, nothing signals hidden strength or weakness better than a stochastic divergence. While the final top in the transports in 1999 did not feature the kind of negative stochastic divergence that most true tops do in fact produce, there was the 2B top/Turtle Soup test to which the transports rapidly succumbed. Interestingly — and I'll have more to say about this later — the industrials did form a negative stochastic divergence on the monthly charts to anticipate its all-time top in December 1999. Does the fact that the transports made their all-time high in 2005 — not 1999 — help explain why there was no negative stochastic divergence in the transports' monthly chart back in 1999?

Figure 2: A positive stochastic divergence in the monthly chart of the Dow transports anticipated a major rally in the average over the following year.

Anyway, there are few clearer bottoms produced by positive stochastic divergences on the monthly scale than the one produced by the Dow transports in late 2002 and early 2003. While the Dow industrials' higher low in early 2003 (vis-à-vis the low in late 2002) was a relatively bullish development in and of itself, the positive stochastic divergence on the transports' monthly chart was the greatest call to bullish arms in the Dow transports since, at least, the mid-1996 correction (which saw the transports fall from 2300 to 2000 in three months before bounding up to nearly 3800 less than two years later). Interestingly, in the spring of 2003, the transports were trading at pretty much the same level then as they were in 1996!

If the transports clearly signaled a bottom in the late 2002-early 2003 period, then they do not appear to be signaling a top in 2005. As the monthly chart of the transports shows, there is precious little potential for a negative stochastic divergence in the transports. The monthly stochastic peak in late 2004-early 2005 is clearly higher than the previous monthly stochastic peak in the second half of 2003. This all but eliminates the possibility of a negative divergence in the monthly transports at this time.

It is worth adding that the transports didn't require a negative stochastic divergence back in 1999 in order to establish a top that would last for the next few years. As a result, traders shouldn't be surprised if the transports in early 2005 were able to once again establish a multiyear top without producing a negative stochastic divergence first. Interestingly, as was the case with the top in 1999-2000, it is the Dow industrials that are doing a better job of warning of a potential top, insofar as a negative stochastic divergence is increasingly prominent in monthly charts of the Dow 30.

Another new high, another new top?

So absence and negative divergence notwithstanding, does anything in the performance of the transports in late 2004 and early 2005 suggest that a top is imminent? After briefly breaking out to new, all-time highs in November 2004, the transports pulled back below those highs in 2005. However, as I've already pointed out, there is no negative stochastic divergence in the transports to make a bearish call on that average all the easier to abide by.

So why not haul out the 1999 toolbox, dust off the monthly 2B top test/Turtle Soup, and apply it to current action in the transports? By this method, it would require a drop to or near the 3400 level before the transports could be said to be in a 2B top. A less-aggressive way of playing a potential 2B top at these levels would be to wait for prices to drop below the level of the low of the month during which the initial top was made (that is, the old "all-time new high" from April 1999). That would suggest something closer to the 3260­3250 level.

It is worth remembering that even if the transports experience significant weakness going forward, the industrials are capable of moving higher before joining the transports in their swoon. Again, in 1999, the transports topped out in April. The industrials did not top until that December — and even then the industrials, unlike the transports, spent the first several months of the bear market moving sideways rather than simply collapsing. Recall that it was a nonconfirmation in 1999 that occurred almost a year before Dow theory confirmed the bear market of 2000­02.

If the much-anticipated (by some) resumption of the bear market is slated to be as powerful and destructive as many suspect it will, then a few months of confusing nonconfirmation before another master-quality Dow theory sell signal should at least be on every trader's and investor's list of possibilities in the coming year.

David Penn may be reached at DPenn@Traders.com.

Suggested reading

Current and past articles from Working Money, The Investors' Magazine, can be found at Working-Money.com.