Similarly, the 1966 stock market peak was preceded by 24 years of basically rising prices followed by a long period of consolidation involving widely swinging stock prices. When adjusted for inflation, stock prices peaked in 1965, subsequently experiencing an extremely severe secular bear market, which was comparable to the 1929–1932 debacle.
Another example comes from the big bull market in gold, which started in 1968 at $32 and ran up to $850 in January 1980. While the price decline was not as severe as the 1929 crash, the next 20 years were spent in a frustrating sideways trading range at prices well under half their peak value. Once that cathartic process had been completed, prices were free to more than double from their 1980 peak.
Major Technical Principle Time is concerned with adjustment because the longer a trend takes to complete, the greater its psychological acceptance and the greater the necessity for prices to move in the opposite direction and adjust accordingly.
Investors become accustomed to rising prices in a bull market, with each reaction being viewed as temporary. When the trend finally has reversed and the first bear market rally takes place, the majority are still convinced that this, too, is a temporary reaction and that the bull trend is being renewed. The initial response is always disbelief, as reflected in the attitude “It’s bound to come back” or “It’s a good company; I am in it for the long-term.” As prices work their way lower in a bear market, the adjustment takes a less optimistic form because the majority of investors forsake their expectations of a rising market and look for them to move sideways for a time. The psychological pendulum finally swings completely to the other (bearish) side, as investors watch prices slip even further and become overly pessimistic. At this point, sufficient time and downside price action have elapsed to complete the adjustment process, and the market in question is then in a position to embark on a new bull cycle.
Time has been viewed here in an emotional context, since it is required for investors to adjust from their unrealized expectations. Both traders and investors should realize that time is deeply bound up with the business cycle. This is due to the fact that a strong and lengthy recovery, like those between 1921–1929 and 1990–2000 breed confidence among investors and businesspeople. As a result they tend to become inefficient, careless, and overextended due to this long period of prosperity. The subsequent contraction in business conditions needed to wipe out these distortions is thus more severe. Equity prices suffer the double influence of: (l) losing their intrinsic value due to the decline in business conditions and (2) being revalued downward from the unrealistically high levels that prevailed during the period of prosperity. The reverse set of circumstances applies following a long market decline.
Major Technical Principle The idea of a reaction commensurate with the previous action is known as the principle of proportionality.
Measuring time as an independent variable is a complicated process, since prices move in periodic fluctuations known as cycles. Cycles can operate for periods ranging from a few days to many decades. At any given moment, a number of cycles are operating simultaneously, and since they are exerting different forces at different times, the interaction of their changing relationships often has the effect of distorting the timing of a particular cycle.
The most dominant of the longer ones is the so-called 4-year cycle, in which there is a nominal or average length between troughs of 41 months. Since several other cycles are operating at the same time but with different influences, the length of the 4-year variety can vary either way by 6 months or so.
Cycles are shown on a chart in the form of a sine wave, as in Figure 24.1. These curves are usually based on a rate of change (ROC) or trend-deviation calculation, which is then smoothed to iron out misleading fluctuations. Since it occurs only rarely that two cycles are of identical length, an average, or nominal, period is calculated. This theoretical time span is used as a basis for forecasting.
FIGURE 24.1 Typical Cycle
Major Technical Principle Since it occurs only rarely that two cycles are of identical length, an average, or nominal, period is calculated. This theoretical time span is used as a basis for forecasting.
In Figure 24.2, this idealized cycle is represented by the dashed line and the actual cycle by the solid one. The arrows indicate the peaks and troughs of the idealized cycle: In actual fact, price trends rarely reverse exactly at theoretical points, especially at peaks, where there is often a long lead time. Nevertheless, the theoretical points provide a useful guide. Three other important principles are concerned with cycle analysis in addition to those of proportionality and nominality. The first is the principle of commonality, which states that cyclicality of similar duration exists in the price action of all stocks, indexes, and markets. This means that a 4-year cycle exists not only for the U.S. stock, bond, and commodity markets, but also for each individual stock and for international markets as well.
FIGURE 24.2 Typical versus Idealized Cycle
Major Technical Principle The greater the number of markets, stocks, or commodities in a universe that are moving in the same direction, the stronger that trend is likely to be.
For example, if two food stocks are experiencing breakouts, the trend for food stocks is likely to be less significant than if, say, 10 stocks are experiencing similar breakouts. The old adage “strength in numbers” certainly come to mind on this point.
Major Technical Principle The principle of variation states that while stocks go through similar cycles, the price magnitudes and durations of these nominal cycles will be different because of fundamental and psychological considerations.
In other words, all stocks, indexes, and markets go through a similar cycle, but the timing of both their peaks and their troughs differs, and so does the size of their price fluctuations. For example, the interest-sensitive and cyclical (basic industry) stocks go through the similar cycle, but because interest-sensitive stocks, such as utilities, lead the market, cyclicals, such as steel groups, generally lag behind them. This is shown in Figure 24.3. Similarly, the interest-sensitive issues may rise by 80 percent from the trough to the peak of their cycle, while the cyclicals might advance by only 20 percent, and vice versa.
FIGURE 24.3 Leading versus Lagging Sectors
Chart 24.1 also illustrates this principle, and shows the interaction of financial series during a typical business cycle. The rising part usually consists of three stages, which correspond to the three phases described in Dow theory. It is normal for prices to reach a new high as each stage unfolds, but sometimes this does not happen. This is known as a magnitude failure, and is a distinct sign of weakness. A magnitude failure occurs because of very poor underlying fundamentals. In effect, the cycle misses a beat. Magnitude failures are a characteristic of a contratrend price movement such as a short-term rally in a primary bear market, a primary trend advance that develops under the context of a secular bear, and so forth.
CHART 24.1 Typical Cycles with Financial Series in Percentage of Their Averages. A Mechanistic Approach to Business Cycle Conditions
The opposite can also occur; exceptionally strong fundamentals (or the perception of them) can give rise to a fourth stage, in which prices undergo an additional upward leg. For equity markets, this final upward surge is often associated with an extended period of declining interest rates. Such strong underlying conditions normally develop when the 4-year cycle occurs in conjunction with the peak of longer-term cycles, such as that associated with secular trends.
In cases in which the cyclic turning points of a number of components of a particular market converge, the magnitude of the next move will be much greater. For example, the turning points of individual stock markets around the world can occur at different times. However, in the summer of 1982, most of their cyclical lows coincided. The resulting rally in virtually all markets was explosive.
The third principle is summation.
Major Technical Principle The principle of summation occurs when several cycles are combined
into the calculation of a specific indicator.
It is really the combination of a number of cycles into one and is the concept behind the Know Sure Thing (KST) market cycle model discussed in Chapter 15. If the result were plotted as one idealized cycle, it would be represented by a curve similar to the Special K indicator, also discussed in the same chapter.
There are four influences affecting a time series trend at any one time: secular, cyclical, seasonal, and random. The cyclical trend is the starting point for the purpose of analyzing primary bull and bear markets. Specifically, this is the 4-year, or Kitchin, cycle. The secular influence is a very long-term one that embraces several 4-year cycles. From the point of view of a stock, bond, or commodity market, the most dominant secular cycle” is ranges from 30 to 50 years between troughs. Two other important cycles in excess of 4 years are the 9.2-year and 18 -year cycles.
Figure 24.4, adapted from Business Cycles by Joseph Schumpeter (McGraw-Hill, 1939), combines the effect of three observable business cycles into one curve. In effect, it shows the summation principle using three longer-term cycles: the 50- to 54-year (Kondratieff), the 9.2-year, and the 41-month (Kitchin) cycles. The model is not intended to be an exact prediction of business conditions and stock prices, but rather to indicate the interaction of the shorter cycles with the longer ones. Even so, it is worth noting that the long-term curve crossed below the zero line in 1987. Projecting the down wave from that point forward to its expected positive equilibrium crossover 20 to 25 years later certainly brings us into the ballpark for the secular low in commodity prices, which actually took place in 2001. A more detailed description of the long wave was presented in Chapter 23. For now, we will concentrate on a few cycles of relatively shorter duration.
FIGURE 24.4a Schumpeter’s Model of Nineteenth-Century Business Cycles. (From Joseph Schumpeter, Business Cycles,McGraw-Hill, New York, 1939.)
FIGURE 24.4b Twentieth-Century Business Cycle and Crisis Points (Calculated Path).
The 18-Year Cycle
Normally, the amplitude of a cycle is a function of its duration; i.e., the longer the cycle, the bigger the swing.
The 18-year or, more accurately, the 18 -year cycle, has occurred fairly reliably in stock market prices since the beginning of the nineteenth century, but its recent performance is questionable. Even so, this cycle gains credibility because it operates in other areas, such as real estate activity, loans and discounts, and financial panics.
The smoothed line in Chart 24.2 is a 3-year centered moving average (MA) of common-stock prices from 1835 to 2012. This average helps to smooth the trend and isolate the long-term picture more clearly. The beginning of the 18-year cycle at major market bottoms is self-evident when the vertical lines are referenced.
CHART 24.2 S&P Composite, 1835–2012 the 18 Year Cycle
While the average cycle lasts 18 years, actual cyclical lows can vary 2 to 3 years either way. The increase in government interference in the economy resulting from the Keynesian revolution has had the effect of biasing the cycle to the upside to the extent that there is a question of whether it is still operating. For example, the conceptual low in the 1987–1988 period did not coincide with a bottom associated with a business cycle. Note that the early 1950s and 1987–1988 lows developed during the course of a secular bull market as defined in the previous chapter. In such cases, we would not expect to find much in the way of downside cyclicality, but those juncture points, like regular cyclic lows, still qualified as a long-term buying opportunity.
On balance, since 1840, the 18-year cycle has operated fairly consistently. However, the erratic nature of the last two cycles leads us to question whether this 18-year periodicity is still continuing to operate, and is another reason why cycles, just like any other technique, need to be augmented with other indicators.
The 17.5-Year Cycle
A slight variation on the 18-year cycle is its 17.5 counterpart. This is featured in Chart 24.3, where you can see that it has embraced many significant bottoms as well as tops. Secular turning points, such as 1929, 1949, 1966, 1982, and 2000, all developed close to expected bottoms or inversions. Unfortunately, this cycle does not distinguish between tops and bottoms—there again, momentum indicators are required for this process.
CHART 24.3 S&P Composite, 1840–2012 the 17.5-Year Cycle
The 9.2-Year Cycle
Chart 24.4 shows the 9.2-year cycle in stock prices from 1830 to 1946. The dashed lines represent the ideal cycle, which reversed exactly on schedule, and the solid line shows the actual annual average as a percentage of its 9-year MA trend.
CHART 24.4 The 9.2-Year Cycle in Stock Prices 1830–1946
The cycle occurred 14 times during the 1930–1946 period, and according to the Bartels test of probability, it could not occur by chance more than once in 5,000 times. Further evidence of the significance of this cycle is given by observation of the 9.2-year periodicity in other phenomena as unrelated as pig iron prices and the thickness of tree rings.
One problem with using the technique illustrated in the chart is that the annual average is expressed as a percentage of a centered 9-year MA. This means that the trend is not known until 4 years after the fact, so there is always a 4-year lag in learning whether the 9.2-year cycle is still operating. Nevertheless, if the theoretical crest in 1965 is used as a base and the 9.2 years are subtracted back to 1919, the peaks of the 9.2-year cycle correspond fairly closely to major stock market tops.
The vertical lines in Chart 24.5 show the 9-year cycle theoretical lows using the 1932 bottom as a centering device. The small black arrows show how the actual low diverged from the theoretical low or became an inverted high. Those against the 55-month ROC (half the time span of the cycle) point up when the cycle turning point corresponded with an important ROC reversal or the start of a more accelerated trend.
CHART 24.5 S&P Composite, 1900–2012 the 9.2-Year Cycle
The cycle worked well until the 1990s—after that, the lows represented good buying points but developed halfway up the rallies. The fact that this cycle completely missed the lows of 2002 and 2009 suggest it may not still be operating.
The Decennial Pattern
This pattern was first noted by Edgar Lawrence Smith in his book Tides and the Affairs of Men (Macmillan, 1932).1 Smith researched equity prices back to 1880 and came to the conclusion that a 10-year pattern, or cycle, of stock price movements had more or less reproduced itself over that 58-year period. He professed no knowledge as to why the 10-year pattern seemed to recur, although he was later able to correlate the decennial stock patterns with rainfall and temperature differentials. Even though the cycle is a relatively reliable one, there has been to date no rational explanation as to why it works.
Smith used the final digit of each year’s date to identify the year in his calculations. The years 1881, 1891, 1901, etc., are the first years; 1882, 1892, etc., are the second; and so forth. Inspired by the research of Dr. Elsworth Huntington and Stanley Jevons,2 who both emphasized the 9- to 10-year periods of recurrence in natural phenomena, Smith experimented by cutting a stock market chart into 10-year segments and placing them above each other for comparison, as shown in Chart 24.6. He concluded from this data that a typical decade consists of three cycles, each lasting approximately 40 months.
CHART 24.6 The Decennial Pattern of Industrial Stock Prices
The late Edson Gould, who came into prominence in the mid-1970s because of his uncannily accurate stock forecasts, used the decennial cycle as a cornerstone for his research. In his 1974 stock market forecast Gould wrote, “In the 35 years that have passed since Mr. Smith’s book was published—35 years of wars, inflation, and vast changes in our economic monetary set-up and background—the action of the stock market has, much of the time, fitted unusually well with the 10-year pattern.”3 Smith’s discovery has stood the test of time.
The stock series in Chart 24.7 represents a simple average of the decennial pattern from 1900 to 2009, giving equal weight
to the proportional movements of each period. It starts off with years ending in zero and ends with those terminating with nine. You can see that the trajectories are very similar during secular bull and bear markets even though they represent different years. The chart also brings out the fact that secular bull years are easier to make money in than secular bears, a fairly obvious observation but nonetheless a very important one.
CHART 24.7 Decennial Cycle 1900–2009 Distinguishing between Secular Bull and Bear Markets
The pattern is plotted together with a 12-month ROC in Chart 24.8. The swings in the ROC show four down phases and an equal number of advances, as flagged by the solid lines.
CHART 24.8 Decennial Cycle 1900–2009 and a Momentum Indicator
The decennial pattern can be of greater value if it is used to identify where the strong and weak points usually occur and then to see whether other technical phenomena are consistent. For example, in the middle of year 9, the 12-month ROC indicator for the average cycle is highly overbought, which is consistent with a decline or consolidation, starting at the end of that year and following through to the year ending in zero. In 1949, the 12-month ROC was highly oversold and was inconsistent with its normal position in the decennial pattern. Instead of declining into 1950, the market actually rose. This experience is a good example of why the decennial approach should be used with other technical indicators and not in isolation.
Important Years
Technical Analysis Explained Page 35
