A sign that a bull market was starting did not come when the MACD reached an overbought reading, which, after all, is still possible in a bear market. Instead, it came when the May decline in the oscillator did not fall back to an oversold reading but was held just above zero. Such action indicated that the underlying character of the MACD had probably changed for the better.
Chart 19.10 shows an analysis of the relative action in greater detail starting with the October top. Signs of weakness started to appear as both the RSI and the MACD of relative strength violated up trendlines. This was then confirmed by the RS line itself violating one of its own. This joint action was not important enough to signal a bear market, but it definitely indicated that the uptrend would likely be stalled for several months. In effect, Abbott Labs was probably underperforming the market during that period. If you look carefully, you can see that the RSI trendline was, in fact, the neckline of a head-and-shoulders top.
CHART 19.10 Abbott Labs RS, 1998–2001 and Short-Term RS Momentum
As we move on, the price action becomes progressively disappointing. The January oversold condition merely triggers a sideways trading range after which the relative downtrend is resumed. Also, look at the three dashed down trendline breaks. They should have been followed by a good rally but they were not—this type of action is often indicative of a bear market.
Chart 19.11 shows the same period but also includes the actual price. The dashed up trendline marks the approximate point where the relative sell signal was triggered, but note that the absolute price continued to extend its rally. It then diverged negatively with the RS line, indicating underlying technical weakness. However, it remained above its solid up trendline until January 1999. If the sell signals in the RS line that had developed previously were not sufficient evidence to justify liquidation, the violation of this trendline in the absolute price certainly was.
CHART 19.11 Abbott Labs RS, 1998–1999 and Short-Term RS Momentum
Chart 19.12 shows a close-up of the bullish period featured in Chart 19.9. Remember, coming into this period the RS line had been in a strong bear market where the momentum indicators had been triggering false signals. However, in March 2000 some positive action by both momentum series starts to develop, since they barely fell below the equilibrium level at the time the RS line was reaching its second low in the top panel. Also, the MACD moved above its previous high, indicating a probable change in character more suitable for a bull than a bear market.
CHART 19.12 Abbott Labs RS, 1999–2000 and Short-Term RS Momentum
Finally, the RS line confirmed by breaking above the horizontal line, marking the top of a double-bottom formation. At the same time, it confirmed that a series of rising peaks and troughs was now under way. Throughout the bear market, each rally high was lower than its predecessor; likewise with the bottoms.
Spreads
RS is widely used in the futures markets under the heading “spread trading,” in which market participants try to take advantage of market distortions. These discrepancies arise because of unusual fundamental developments that temporarily affect normal relationships. Spreads are often calculated by subtracting the numerator from the denominator rather than dividing. I prefer the division method because it presents the idea of proportionality. However, if a spread is calculated over a relatively short period (for example, less than 6 months), it is not important whether subtraction or division is used.
Spread relationships arise because of six principal factors:
• Product relationships, such as soybeans versus soybean oil or meal, oil versus gasoline or heating oil
• Usage, such as hogs, cattle, or broilers to corn
• Substitutes, such as wheat versus corn or cattle versus hogs
• Geographic factors, such as copper in London versus copper in New York or sugar in Canada versus sugar in New York
• Carrying cost, such as when a specific delivery month is out of line with the rest
• Quality spreads, such as T-bills versus eurodollars or S&P versus Value Line
Some of these relative relationships, such as London versus New York copper, really represent arbitrage activity and are not suitable for the individual investor or trader.
On the other hand, the so-called TED Spread, which measures the relationship between high-quality T-bills and lower-quality eurodollars, is a popular trading vehicle.
In some cases, spreads move to what was previously an extreme and proceed to an even greater distortion. For this reason, it is always important to wait for some kind of trend-reversal signal before taking a position. While the risk associated with such transactions is by no means eliminated, it will certainly be reduced.
Other relationships between various asset categories are further analyzed in subsequent chapters. These relationships may be used for different purposes, but all are subject to trend reversals that can be identified by the techniques already described.
Summary
1. Comparative relative strength compares one security with another. The result is plotted as a continuous line.
2. The most common application is to compare a stock or sector with a market average. When the line is rising, it means that the security in question is outperforming the market and vice versa.
3. Divergences between the absolute price and relative strength warn of latent strengths and weaknesses and can help in identifying potential trend reversals.
4. Relative strength moves in trends. Any legitimate trend-determining technique can be applied to a relative line.
5. Joint trendline breaks in both the absolute price and the RS line usually result in reliable trend-reversal signals.
6. One of the most useful techniques for analyzing the primary trend of relative action is the use of smoothed long-term oscillators, especially the KST.
20 PUTTING THE INDICATORS TOGETHER: THE DJ TRANSPORTS 1990–2001
It is now time to combine the indicators that we have covered so far into an analysis of the long-term picture. For this purpose I’ve chosen the Dow Jones transportation average between 1990 and 2001. Chart 20.1 shows the average together with its 9-month MA. This was one of the best testing averages derived by optimizing from 1931 through to the year 2000.
CHART 20.1 Dow Jones Transports, 1989–2001, and Turning Points
The upward and downward pointing arrows indicate the principal turning points in this period. The 1990 bottom was not an easy one to recognize because the average virtually reversed on a dime. Chart 20.2 shows that the 18-month rate of change (ROC) violated a sharp down trendline just before the price.
CHART 20.2 Dow Jones Transports, 1989—2001, and Long-Term Momentum
The relative strength (RS) line, in the center panel of Chart 20.3 actually broke its bear market trendline ahead of the absolute price. This indicated that the DJ Transports were likely to outperform the market during the early stages of the new bull market.
CHART 20.3 Dow Jones Transports, 1989—2001, and Relative Strength
The vertical line in Chart 20.4 shows that this was one of the few occasions in which all three oscillators were simultaneously oversold. This chart also offered the strongest buy signal because the down trendline for the price was violated at approximately the same time as the 65-week exponential moving average (EMA). Also, the 39-week CMO completed a base. By February 1991, several positive signs had therefore developed, all of which indicated that downside momentum had probably dissipated sufficiently to allow the long-term know sure things (KSTs) to reverse to the upside.
CHART 20.4 Dow Jones Transports, 1989—2001, and Three Weekly CMOs
The next major event was the intermediate peak in 1992. The average briefly crossed below its 12-month (Chart 20.2) and 65-week EMAs (Chart 20.3), and the long-term KST also triggered a negative whipsaw signal. These events could certainly have justified the conclusion that the Transports had begun a bear market. However, once the average and the long-term KST (Chart 20.2) had crossed back above their r
espective moving averages (MAs), there was little reason to maintain a bearish stance.
Unfortunately, this whipsaw type activity occasionally develops from an intermediate correction. Under such circumstances it is important to keep an open mind on the indicators. In this case, Chart 20.4 shows that the 20-week Chande momentum oscillator (CMO) broke out from a base and several down trendlines were broken, so there was plenty of evidence that the tide had turned.
The top of the bull market developed 2 years later in early 1994. Signs of a major top were quite widespread. In Chart 20.2 the Transports simultaneously violated a 4-year up trendline and the 12-month MA. The KST gave a decisive sell signal, and the 18-month ROC completed a head-and-shoulders (H&S) top. In the whole 11 years covered by the chart, there were only two completed chart patterns for this indicator, so the early 1994 breakdown was very significant.
Chart 20.4 was equally significant in its bearish entrails. The 39- and 52-week CMOs diverged negatively with the price, and both series completed a top or experienced a major trendline violation. At the peak itself, the 20-week CMO was actually overbought. The chart shows that, except for the strongest of up- or downtrends, the overbought and oversold conditions were often associated with intermediate-type reversals. Later in the year something more ominous started to happen, and this was a trendline break in the RS line. For the first time since the bull market in RS began the long-term RS, KST (shown in Chart 20.3) triggered a decisive sell signal. Although it was not apparent at this point, the transports had begun a long period of underperformance.
Because the ensuing bear market was relatively mild, the bottom in early 1995 was only signaled on the weekly charts. Once again, Chart 20.4 holds the key, as the down trendline in the 52-week CMO was violated and a base in the 20-week series completed. The Transports themselves more or less simultaneously broke above their bear market down trendline and 65-week EMA. The long-term KST in Chart 20.3 also turned positive around the same time.
The average remained above its 65-week EMA for the next 4 years and the series of rising peaks and troughs continued. Then, some extremely serious trend breaks developed. First, the average itself crossed below its 12-month MA and violated its bull market trendline (Chart 20.2). The KST also triggered a sell signal and the 18-month ROC completed a top.
Chart 20.5 shows that the Transports also completed and broke down from an upward-sloping H&S top and crossed below its 65-week EMA more or less at the same time. Notice how the 39- and 52-week CMOs were actually below zero at the time the average was forming the right shoulder. This distinct lack of upside momentum was a very bearish sign. Not surprisingly, the Transports experienced a pretty sharp decline into the fall of 1998.
CHART 20.5 Dow Jones Transports, 1995—2001, and Three Weekly CMOs
The most serious technical damage of all came from Chart 20.3 in the form of a major breakdown in the RS line below a 6-year support trendline. This happened as the absolute price was crossing below its 65-week EMA. Trouble on the RS front had been signaled long before this, because it had utterly and completely failed to confirm the bull market in the absolute price. By the time of the 1998 peak in the Transports themselves, the RS line had experienced a major negative divergence. When the RS line moved to a new post-1993 low at the end of 1996, this should have been warning enough that there were far better places for exposure than transportation stocks.
The 1998 bottom, like that of 1990, was an elusive affair, but more so because the turn was so sharp. All three CMOs in Chart 20.5 violated down trendlines, but the average itself did not cross above its average until it had rallied a long way from the bottom. No down trendlines could be drawn against the price, so it was not really possible to build a timely and convincing bullish case. In cases where the evidence of a trend reversal is incomplete, it is always better to avoid the security in question. In any event, the overriding factor should have been the early 1998 breakdown in the RS line, as this set the scene for the next several years of trading action.
Indeed, as it turned out, the 1998–1999 advance was really an above-average reflex rally since all the price was able to do was to rally back to resistance in the form of its 1998 high and the extended bull market trendline. During this whole period the KST failed to give a buy signal and the RS line in Chart 20.3 never crossed back above its 65-week EMA.
Finally, the average violated its 1990–2000 up trendline at the start of the new century. This was not a great trendline, for, although it was quite long, it had only been touched on two occasions and was not therefore a great reflection of the underlying trend. However, it did result in a sideways trading range over the ensuing 2 years. The critical point after this would be the trendline joining a series of lows between 1996 and 2001. A break below it, not with standing other evidence, would be a serious technical blow.
Summary
This has been a brief account of the technical position of the Transports between 1990 and 2000. Although it was not possible to include too many indicators, it has enabled us to describe how trend indicators for price, momentum, and RS can be combined to help identify major turning points.
Part II
MARKET STRUCTURE
21PRICE: THE MAJOR AVERAGES
In previous editions, our central theme focused on technical principles, with a primary objective of analyzing the U.S. stock market. In the early 1980s, when the United States was far more dominant in the global financial scene, that approach had some merit. In the second decade of the twenty-first century and beyond, the attention of technicians has become far more diverse, having broadened to international stock markets, bonds, commodities, and currencies. While our coverage will be broadened in this edition, it is not possible to cover all of the market averages and indexes that have been developed for these various entities in one small chapter.
Another important financial market development in the twenty-first century has been the rapid expansion and burgeoning popularity of exchange-traded funds (ETFs) and, to a lesser extent, exchange-traded notes (ETNs). Previously, the purchase of an index involved the acquisition of its individual components, but with ETFs, it was now possible to buy the index just like a stock. That’s because the ETF is a basket of stocks whose management objective is to replicate a stated index. To give you an idea of their growth, I stated in my book The Investor’s Guide to Active Asset Allocation (McGraw-Hill, 2006) that there were then 160 listed ETFs. In mid-2013, that number was closer to 1,500 and still growing. One personally important introduction was the Pring Turner Business Cycle ETF, in December 2012 (symbol DBIZ). The fund adopts an active approach based on the business cycle and technical strategies outlined in this book.
In this chapter we are going to examine some of the principal U.S. indexes and expand the discussion to cover some of the ETFs that reflect major global equity, bond, and commodity indexes.
U.S. Equities
There is no ideal index that represents the movement of “the market.” It’s true that the majority of stocks move together in the same direction most of the time, but there is rarely a period when specific stocks or several industry groups are not moving contrary to the general direction of the trend. The general level of stock prices is basically measured in two principal ways. The first, known as an unweighted index, takes a mean average of the prices of a wide base of stocks; the second also takes an average of the prices of a number of stocks, but in this case, they are weighted by the capitalization of each company (i.e., the number of shares outstanding multiplied by their price). The first method monitors the movement of the vast majority of listed stocks, but since the second gives a greater weight to larger companies, movements in a market average constructed in this way more fairly represent changes in the value of investor portfolios. For this reason, weighted averages are usually used as the best proxy for “the market.” These averages are compiled from stocks representing public participation, market leadership, and industry importance.
Several price indexes have be
en developed that measure various segments of the market. Their interrelationship offers useful clues about the market’s overall technical condition. Chapter 3 discussed in detail the relationship between the Dow Jones Industrial Average (DJIA) and the Dow Jones Transportation Average, but there are many other useful indexes, such as the Dow Jones Utility Average, unweighted indexes, and a few bellwether stocks groups. They are examined in this chapter in the context of their contribution to the U.S. market’s overall technical structure.
Composite Market Indexes
The DJIA is the most widely followed stock market index in the world. It is a price-weighted average and is constructed by totaling the prices of 30 stocks and dividing the total by a divisor. The divisor, which is published regularly in The Wall Street Journal and Barron’s, is changed from time to time because of stock splits, stock dividends, and changes in the composition of the average. In recent decades, its makeup has expanded from its industrial base to include consumer goods financials and other sectors. However, strictly speaking, it is not a “composite” index, since it does not include such industries as transportation or utilities. Yet, the capitalization of the DJIA is still equivalent to a substantial percentage of the outstanding capitalization on the New York Stock Exchange (NYSE), and it has normally proved to be a reliable indicator of general market movements. The original reason for including a relatively small number of stocks in an average was convenience. Years ago, the averages had to be laboriously calculated by hand. With the advent of the computer, the inclusion of a more comprehensive sample became much easier.
Technical Analysis Explained Page 29