Dow Theory: Confirmations 2004

This means that the Dow industrials and transports tend to trend together over the long term. When these averages are trading "in gear" with each other, Dow theorists say that the averages are confirming each other. An excellent example of this confirmation came during the bull market period in late 2003, when the industrials and transports often made new highs on the same day.

Other times, however, the averages do not trade together. In such instances, one average may go on to set new highs, while the other average remains flat or even sinks to new lows. Because Dow theory holds that "The Averages Must Confirm," Dow theorists encountering divergences between the Dow industrials and transports can say only that a "nonconfirmation" exists. The previous Dow theory bullish or bearish signal should be considered still intact until proven otherwise. Moreover, Dow theorists expect that the "nonconfirmation" will eventually be worked off, with one average abandoning its bullish or bearish course to once again trade in-gear with the other.

Call me nonconfirmed

In Technical Analysis Of Stock Trends, Dow theorists Robert Edwards and John Magee note of the principle that "The Averages Must Confirm":

It is not necessary that the two averages confirm on the same day. Frequently both will move into new high (or low) ground together, but there are plenty of cases in which one or the other lags behind for days, weeks or even a month or two. One must be patient in those doubtful cases and wait until the market declares itself in definite fashion.

Nevertheless, we can't help but wonder if Edwards and Magee would have been able to maintain their patience in the face of the massive nonconfirmation between the Dow industrials and Dow transports that has existed since early June 2004. Who among us would have believed it was a mere nonconfirmation that took place when the Dow transports set a new higher high in June (vis-à-vis the April highs) while the Dow industrials failed to do so? Not only did the transports set a higher high, but its June peak actually set a new high for the year!

But the market gods had more consternation in store for Dow theorists than a single nonconfirmation. In August, the Dow industrials completed a lower low, while the Dow transports formed a higher low (higher vis-à-vis the May lows). And by September, while the Dow industrials were marking off their third consecutive lower peak in 2004, the Dow transports were actually setting still another year-to-date high.

Powerfully bullish action on the part of the transports should not be a surprise to long-term chart watchers. In the months following the crash of 1987, the Dow industrials went from a low of 1600 to a high of 3025 before topping out and falling into the bear market of the early 1990s. That was a move of 89%. At the same time — or, to be accurate, in less time — the transports climbed from 650 to 1500, a 131% gain. The industrials made their move from October 1987 to July 1990. The transports began in December 1987, but topped out in August 1989, almost a year before the industrials did.

Nevertheless, the exuberant bullishness in the transports has complicated the analysis. While the Dow industrials have yet to set a higher low for the year (in fact, the industrials have not had a significant correction since the average bottomed in late October 2004), there are a number of signs that the correction of 2004 may indeed be over. And it was during a rereading of one of my favorite contemporary writers on Dow theory, Vic Sperandeo, that I came across what might be the most instructive and most applicable of these signs. What he says about the current market environment could be as important for despairing bears as for exuberant bulls.

Primaries and secondaries

Writing about Dow theory in his book, Methods Of A Wall Street Master, Victor Sperandeo reminds us of the key observations on primary and secondary market moves made by Robert Rhea. Sperandeo distilled these observations into five "theorems" that describe the nature of primary bull and bear markets, and the secondary reactions or secondary corrections that occur during them.

While Sperandeo's "theorems" deserve to be read in their entirety (I typed them out and keep a copy by my monitor — even if I don't consult them as frequently as I should), I think the conditions that his theorems point to can be summarized as the ability of the averages to pass four main confirmations or tests.

The price confirmation. This is the most obvious and best-known of the Dow theory tests, and states simply that the averages must take out a prior significant high or low. The averages don't have to do this simultaneously. But as I have written in previous Dow theory articles, simultaneous price confirmation between the industrials and transports is not at all uncommon.

The first price confirmation of 2004 was a bearish one, as I've pointed out elsewhere. For the Dow industrials, which were in a primary bull market ever since the October 2002 lows, the relevant significant low is the correction low on January 28 right before the market made a top. This low was exceeded in the first half of March. As for the Dow transports, their prepeak correction low arrived on November 21, and that correction low was taken out in early February.

This was hardly simultaneous confirmation — especially in comparison with the later stages of the October 2002-January 2004 primary bull market, during which same-day confirmations were frequent. But as far as the tenets of Dow theory price confirmation are concerned, in the first few months of 2004, the averages did confirm — and to the downside.

The time confirmation. To a large degree, the use of time as a consideration in Dow theory is Sperandeo's personal contribution to the analysis. Even if it isn't considered a part of "classic" Dow theory, the relative duration of various market moves can be a helpful factor in deciding when primary moves have ended and whether a secondary reaction represents a new primary market.

One of the more interesting discussions on the role of time in technical analysis and trading can be found in Mark Fisher's book, The Logical Trader:

The vast majority of traders I know trade on price, but not on time. How many people have you heard say that if you take a position and it doesn't go anywhere in 20 or 30 minutes, then you should get out? Very few, if any. I'm here to state that time is the most important factor in trading. If the scenario you've envisioned doesn't materialize within a certain time frame, then just move on and look for the next trade.
How do you factor time into your trades? Simple. You set the time parameters for a certain scenario to occur.

How does this relate to Dow theory? Sperandeo suggests that a primary market must last for at least six months. Anything clearly shorter than that is likely a secondary correction rather than a new primary market. This kind of analysis has allowed Sperandeo to acknowledge bear markets while others were still clinging to hopes of a resurgent bull, as well as to note certain events like the crash of 1987 as being quite a roller-coaster ride, but not representative of a true bear market.

Here is where the confirmation gets murky. Even given the current rally in the Dow industrials, the pattern of lower highs and lower lows that began in January 2004 and continued through October remains intact. At this point, this would be true even if the current rally turns out to be the beginning of a new primary bull market (rather than a resumption of the primary bull from October 2002). So the Dow industrials pass the time confirmation test.

However, the Dow transports clearly do not. Topping in January 2004, the transports made a low in March and then a higher low in May. Top to bottom, the bear market in the transports lasted less than three months. That suggests that the downward movement in the transports in 2004 was more indicative of a secondary reaction in an ongoing primary bull market than of a primary bear market at the emerging end of a primary bull market.

The magnitude confirmation. Typically, mainstream market watchers suggest that a bear market isn't a bear market until prices decline at least 20% from the top. There's nothing sacred about the 20% level. But it suggests simply that a decline must be significant in degree to warrant the label "bear market." For Sperandeo's part, he suggests in his interview with Jack Schwager in The New Market Wizards that 15% is a minimum correction before a trader or investor considers a correction to be the beginning of a new primary move. Another way of looking at bear market magnitude is to see how much of the previous rally the bear market corrected or retraced. Fibonacci retracement levels can come in handy as guideposts here.

Figure 1: How low can you go? Apparently not low enough for a bear market confirmation based on the magnitude of the correction. The 10% decline from the January 2004 was not enough, in and of itself, to confirm a bear market.

By this criterion, what the price confirmation giveth, the magnitude confirmation taketh away. While the Dow industrials retraced close to 30% of the primary bull market from October 2002 to January 2004, they fell only 10% from the January 2004 peak, falling significantly short of Sperandeo's 15% mark. The case for the Dow transports was similar: a 29% retracement of the average's primary bull market (the transports bottomed in February 2003, as opposed to October 2002), but an 11% drop from the January 2004 peak. Thus, based on the rules of the magnitude confirmation, both averages were close to confirming bear markets, but didn't quite do it.

The bull market confirmation. The last confirmation that is relevant here is, in some ways, the most obvious: If the market makes moves consistent with a bull market (per Dow theory), then it is likely that a bull market is under way. What are "moves consistent with a bull market"? Essentially, this is the same thing as the price confirmation — only in reverse. To wit, if the market during its correction or primary move closes above the high of the last significant bear market peak (that is, bear market rally high) and shows followthrough, then the market can be said to be back in bull mode.

Figure 2: The summer breakout in the transports was a bullish confirmation as long as the rally held. After a 45-day pullback, the rally resumed and the transports' "Correction 2004" was over.

While the Dow industrials were making a new low in August, the Dow transports were making a new high. More specifically, the transports had exceeded the high from January 2004. Having corrected back below that level and rallied back above and beyond that bull market confirmation level, the transports declared as clearly and as loudly as possible that their 2004 correction was over. Again, the key high to overcome for the transports was the previous bear market high — or the high before the last/lowest trough in the correction. With the correction low occurring in March 2004 in the transports, that previous high could only be the January highs.

As of this writing, the Dow industrials have failed to take out their January highs. But focusing on the specific rule that prices must exceed the previous bear market rally peak suggests that the industrials notched a bull market confirmation in early November when the September highs (the highs that immediately preceded the October lows) were bested. This would be months after the bull market confirmation in the transports (which made a bull market confirmation in July). That is stretching the notion of confirmation lagging by "a month or two," but the apparent strength of the move up since late October is such that those trying to fade a bullish confirmation between the industrials and transports — however distant in time — most likely do so at their peril.

Where does this leave us? The bear market price confirmation from January appears to have been trumped by bull market price confirmations in July and November. Neither the transports nor the industrials met the bear market magnitude confirmation test, and only the Dow industrials met the time confirmation test. These "test results" do not specifically point to a bullish or bearish environment going forward. But clearly the burden of believing what you are seeing is on the shoulders of those who simply refuse to believe what they see. And what anyone who has been paying attention has seen is a market geared to go up. As Victor Sperandeo wrote:

The reasons that John Trader, or anyone else in the real world, acts in defiance of knowledge are very complex ... But if I had to reduce the reason to a single sentence it would be: He was trying to avoid the pain of being wrong. It wasn't the money loss that threatened him so deeply — a good trader accepts losing money on some trades as part of the business. What brought about the overwhelming fear that made him behave irrationally was the possibility of suffering the pain of being wrong. He acted to avoid pain, and therein was his problem.

Often the clues to market behavior that analyses like Dow theory can provide will find a trader at odds with preconceived notions of how the market should behave. The advantage of Dow theory — with its effective application by traders over a span of decades — is not that it will make traders forever immune to the "pain of being wrong." Rather, a sound application of Dow theory — as with other tried-and-true technical methods — will help a trader know if and when he or she is wrong and, through a counterconfirmation process, exactly what to do to make things right.

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

Suggested reading

Fisher, Mark B. [2002]. The Logical Trader, John Wiley & Sons.
Schwager, Jack [1992]. "Markets Grow Old Too," interview with Victor Sperandeo, The New Market Wizards, John Wiley & Sons.
Sperandeo, Victor [1991]. Trader Vic — Methods of a Wall Street Master, John Wiley & Sons.
_____ [1994]. Trader Vic II — Principles Of Professional Speculation, John Wiley & Sons.
Charts courtesy Prophet Financial Systems

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