Technical Analysis Explained Page 22

  Hinge When either the %K line or the %D line experiences a slowdown in velocity, indicated by a flattening line, the indication is usually that a reversal will take place in the next period. Examples are shown in Figure 14.7 and Chart 14.24 in the form of a 10-minute chart for Pfizer.

  FIGURE 14.7 Stochastic Hinges

  CHART 14.24 Pfizer and Two Hinges

  Slowing It Down for Better Signals

  It is often a useful idea to tweak the parameters to obtain a slower-moving indicator where %K and %D crossovers can be used to good effect. Chart 14.25, featuring the S&P Composite, shows a 20/20/20 and a 20/10/10 combination. Both show divergences, but the smoother series experiences just one whipsaw, whereas the 20/10/10 variety experiences three. The trade-off is that the more volatile series triggers crossovers on a timelier basis.

  CHART 14.25 S&P composite, 2010–2011 Negative Divergences

  Finally, we move up to the monthly charts where the 24/15/10 combination in Chart 14.26 really comes into its own. This combination, in most cases, will reflect primary trend swings in a fairly good way. For example, there is a positive stochastic crossover at the end of the 2000–2002 bear market confirmed with a nice trendline violation. There was another buy signal confirmed in 2006. Finally, a sell signal in late 2007 was essentially confirmed simultaneously with a trendline penetration.

  CHART 14.26 Echostar, 1998–2011 Stochastic with Long-Term crossover Signals

  Summary

  1. The RSI is bounded by 0 and 100. Overbought and oversold lines should be wider as the time span gets shorter.

  2. The RSI lends itself more easily to comparing the momentum of different securities than does the ROC.

  3. The RSI can be used in conjunction with overbought and oversold lines, divergences, price patterns, trendlines, and smoothings.

  4. Trend-deviation indicators are calculated by dividing the close or a short-term MA by a longer-term one.

  5. Trend-deviation indicators can be used with trendlines, price patterns, and MAs. They also lend themselves to overbought/oversold and divergence analysis.

  6. The MACD is a form of trend-deviation indicator.

  7. The stochastic indicator assumes that prices close near the low at the end of a rally and near their highs at the end of a downtrend.

  8. The stochastic indicator is confined between 0 and 100 and consists of two lines, the %K and the %D.

  9. Stochastic indicators lend themselves to crossovers, divergences, hinges, extremes, and reverse divergences and are usually plotted in their slowed version.

  15 MOMENTUM III: INDIVIDUAL INDICATORS

  Summed ROC: “Know Sure Thing” (KST)

  The Long-Term KST

  We are going to start off this KST explanation by discussing an indicator suitable for price swings associated with primary trends or those that revolve around the so-called 4-year business cycle, subsequently turning to its application for intermediate- and short-term trends.

  Chapter 13 explained that rate of change (ROC) measures the speed of an advance or decline over a specific time span, and is calculated by dividing the price in the current period by the price N periods ago. The longer the time span under consideration, the greater the significance of the trend being measured. Movements in a 10-day ROC are far less meaningful than those calculated over a 12- or 24-month time span and so forth.

  The use of an ROC indicator helps to explain some of the cyclical movements in markets, often giving advance warning of a reversal in the prevailing trend, but a specific time frame used in an ROC calculation reflects only one cycle. If that particular cycle is not operating, is dominated by another one, or is influenced by a combination of cycles, it will be of little value.

  Major Technical Principle At any one time, price is determined by the interaction of many different time cycles. An indicator that takes this into consideration is likely to be timelier without losing too much in the way of sensitivity.

  This point is illustrated in Chart 15.1, which shows three ROC indicators of different time spans: 9 months, 12 months, and 24 months.

  CHART 15.1 S&P Composite, 1978–1988 Three Rates of Change

  The 9-month ROC tends to reflect all of the intermediate moves, and the 24-month series sets the scene for the major swings. The arrows flag the important turning points in this period. They show that, for the most part, all three ROCs are moving in the same direction once the new trend gets under way. A major exception occurred at the 1984 bottom. Here we see the price rise, but immediately after, the 24-month ROC declines while the others continue on up. During the period covered by arrow A, the speed of the advance is curtailed because of the conflict between the three cycles. Later on, though, all three ROCs get back in gear on the upside, and the rally approximated by arrow B is much steeper. In effect, major turning points tend to occur when several cycles are in agreement, and speedy advances and declines develop when more cycles are operating in the same direction. Even this is a fairly limited view because there are far more than three cycles operating at any one point in time.

  Clearly, one ROC time span taken on its own does not give us a complete picture. This was one of the factors considered in the design phase of the KST. Another requirement was an indicator that fairly closely reflected the major price swings over the time period under consideration, primary trends for monthly charts, short-term trends for daily charts, and so forth.

  Chart 15.2 shows the S&P during the 1980–2012 period.

  CHART 15.2 S&P Composite, 1980–2012 and a Smoothed 24-Month ROC

  The oscillator is a 24-month ROC smoothed with a 9-month MA. This series certainly reflects all of the primary trend swings during this period. However, if we use the indicator’s changes in direction as signals, close examination shows that there is a lot to be desired. For example, the 1984 low is signaled with a peak in the oscillator. Similarly, the 1989 bottom in the momentum series develops almost at the rally peak. Also, the 1998 low was signaled with a peak in the oscillator, and the 2005 momentum peak was clearly premature. What is needed, then, is an indicator that reflects the major trend, yet is sensitive enough to reverse fairly closely to the turning points in the price. We are never going to achieve perfection in this task, but a good way of moving toward these goals is to construct an indicator that includes several ROCs of differing time spans. The function of longer time frames is to reflect the primary swings, while the inclusion of the shorter ones helps to speed up the turning points. The formula for the KST is as follows:

  TABLE 15.1 Formula for Time Frames Used in the Long-Term KST

  Since the most important thing is for the indicator to reflect the primary swings, the formula is weighted so that the longer, more dominant time spans have a larger influence.

  Chart 15.3 compares the performance of the smoothed 24-month ROC to the long-term KST between 1974 and 1990. It is fairly self-evident that the KST reflects all of the major swings being experienced by the smoothed 24-month ROC. However, the KST turning points develop sooner than those of the ROC. The vertical arrows slice through the ROC as it bottoms out. In every instance, the KST turns ahead of the arrow, the lead time varying with each particular cycle. Note how in 1988 the KST turns well after the 1987 bottom, but just at the time when the market begins to take off on the upside. The ROC reverses direction much later. There is one period when the KST underperformed, and that is contained within the 1986–1987 ellipse where the KST gave a false signal of weakness, unlike the ROC, which continued on up.

  CHART 15.3 S&P Composite, 1974–1991 The KST Compared to a 24-Month Smoothed ROC

  The dominant time frame in the KST’s construction is a 24-month period, which is half of the so-called 4-year business cycle. This means that it will work best when the security in question is experiencing a primary uptrend and downtrend based on the business cycle. For example, Chart 15.4 shows the KST during the 1960s and 1970s, where the S&P was in a clearly defined business cycle–type trading range. Periods of accumulation and distribu
tion occur between the time when the KST and its moving average (MA) change direction. There are really three levels of signaling. The first occurs when the indicator itself changes direction, the second when it crosses its 9-month MA, and the third when the MA itself reverses direction. In most cases, the MA crossover offers the best combination of timely signals with a minimum of whipsaws. Changes in the direction of the 9-month MA offer the most reliable signals, but these usually develop well after the turning point. The timeliest and most reliable signals thereby develop in those situations where the MA reverses close to a turning point. If it does not, this event should merely be interpreted as a confirmation of a move that is already in progress.

  CHART 15.4 S&P Composite, 1963–1979 Long-Term KST in a Volatile Cyclical Environment

  For the most part, the KST has been reasonably reliable, but like any other technical approach, it is by no means perfect. For instance, the same calculation is shown in Chart 15.5, but this time, for the Nikkei. During periods of a secular or linear uptrend (as occurred for Japanese equities in the 1970s and 1980s), this type of approach is counterproductive since many false bear signals are triggered. However, in the vast majority of situations, prices do not experience such linear trends, but are sensitive to the business cycle. It is for this reason that I call this indicator the KST. The letters stand for “know sure thing” (KST). Most of the time, the indicator is reliable, but you “know” that it’s not a “sure thing.”

  CHART 15.5 Nikkei, 1975–1992 KST Operating in a Linear Trend Environment

  Major Technical Principle The KST should always be used in conjunction with other indicators.

  The principles of interpreting the long-term KST are the same as any other oscillator, though its “default” technique is to observe positive and negative MA crossovers. Occasionally, it’s possible to construct trendlines and even observe price patterns, as well as conduct overbought/oversold analysis.

  Short- and Intermediate-Term KSTs

  The KST concept was originally derived for long-term trends, but the idea of four smoothed and summed ROCs can just as easily be applied to short-term, intermediate, and even intraday price swings. Formulas for various time frames are presented in Table 15.2. These are by no means the last word and are suggested merely as good starting points for further analysis. Readers may experiment with different formulas for any of the time frames and may well come up with superior results. When experimenting, strive for consistency, never perfection, for there is no such thing in technical analysis.

  TABLE 15.2 Suggested KST Formulas*

  Chart 15.6 shows an intermediate KST for the U.S. Dollar Index. The dark highlights indicate when the KST is above its 10-week exponential moving average (EMA), and the lighter plots when it is below. Two round-trip signals have been flagged with the arrows to demonstrate this. The principle of countercyclical trends being weak and the danger of trading them can also be appreciated from this chart. The three small arrows in the 2007–2008 period indicate bullish signals that ran counter to the main trend, which was negative. All three were short-lived and lost money.

  Chart 15.6 U.S. Dollar Index, 2006–2012 and an Intermediate KST

  Chart 15.7 features the London copper price. One of the things I like about the indicator, especially in its short-term and intermediate varieties, is its flexibility of interpretation. Nearly all of the interpretive techniques discussed in Chapter 13 can be applied. In Chart 15.7, for example, we see an overbought crossover at the tail end of 1994. It did not amount to anything because it was not possible to come up with any trend-reversal signals in the price. Later on, though, in early 1995, we see a negative divergence, and at the end of the year an overbought crossover, a 65-week EMA crossover, and a head-and-shoulders top in the price—classic stuff. There were a couple of false buy signals on the way down, but the rally peaks in the KST lent themselves to the construction of a nice down trendline. The violation of the line, the oversold crossover, and the completion of the base in the price combined together to offer a nice buy signal in late 1996. The next time the KST crossed its oversold level in 1998, there was no good place to observe a trend-reversal signal in the price. That was not true in early 1999, where a positive divergence, EMA crossover by the KST, and a trendline break in the price offered a good timely entry point. Note that after the price broke above the down trendline it subsequently found support at the extended line.

  Chart 15.7 Cash Copper, 1993–2001 Intermediate KST Interpretation

  Chart 15.8 features the Global X FTSE Columbia ETF with a daily KST. It demonstrates the fact that this series has the occasional ability to trace out price patterns, as it does in late 2010. Note also the bear market that began late in that year and the inability of the KST to rally much above the equilibrium level.

  CHART 15.8 Global X Columbia ETF, 2009–2012 and a Daily Short-Term KST

  Major Technical Principle Sometimes, a study of the characteristics of an oscillator can be helpful in identifying the direction of the primary trend. Weak rallies indicate bear markets, and weak reactions indicate primary uptrends.

  Note also the ability of the KST to signal divergences, such as the negative discrepancy in early 2012. Chart 15.9 shows the same information as its predecessor, but this time the solid arrows flag positive and negative overbought/oversold crossovers that were confirmed by the price. The failed (dashed) arrows show the value of waiting for such signals to be confirmed by the price.

  CHART 15.9 Global X Columbia ETF, 2009–2012 Daily Short-Term KST Interpretation

  Major Technical Principle The first rule in using the short-term KST is to try to get a fix on the direction and maturity of the primary trend and never trade against it.

  Identifying the direction of the main trend is easier said than done. However, if you pay attention to the price relative to its 65-week or 12-month MA and the level and direction of the long-term KST, or Special K (explained later in this chapter), you will at least have an objective measure of the primary trend environment.

  Chart 15.10, featuring Indian Hotels, again brings out the concept of constructing overbought and oversold lines and seeing what happens when the KST reverses from such levels. In this exercise, I am not concentrating on confirmations, but in qualifying the distinction between the generation of pro- and contratrend signals. The solid arrows indicate pro-trend signals, and the dashed ones indicate contratrend signals.

  CHART 15.10 Indian Hotels, 1998–2003 and a Weekly Short-Term KST

  Using the KST in the Market Cycle Model

  Three Main Trends Earlier chapters explained that there are several trends operating in the market at any particular time. They range from intraday, hourly trends right through to very long-term or secular trends that evolve over a 19- or 30-year period. For investment purposes, the most widely recognized trends are short-term, intermediate-term, and long-term. Short-term trends are usually monitored with daily prices, intermediate-term with weekly prices, and long-term with monthly prices. A hypothetical bell-shaped curve incorporating all three trends is shown in Figure 1.1.

  From an investment point of view, it is important to understand the direction of the main, or primary, trend. This makes it possible to gain some perspective on the current position of the overall cycle. The construction of a long-term KST is a useful starting point from which to identify major market cycle junctures. The introduction of short-term and intermediate series then allows us to replicate the market cycle model.

  The best investments are made when the primary trend is in a rising mode and the intermediate and short-term market movements are bottoming out. During a primary bear market, the best selling opportunities occur when intermediate and short-term trends are peaking and the long-term series is declining.

  In a sense, any investments made during the early and middle stages of a bull market are bailed out by the fact that the primary trend is rising, whereas investors have to be much more agile during a bear market in order to capitalize on the rising inte
rmediate-term swings.

  Combining the Three Trends Ideally, it would be very helpful to track the KST for monthly, weekly, and daily data on the same chart, but plotting constraints do not easily permit this. It is possible, though, to simulate these three trends by using different time spans based on weekly data, shown for the S&P European 350 ETF in Chart 15.11.

  CHART 15.11 S&P Europe ETF 350, 2006–2011 and Three KSTs

  Note that the formula for the short-term KST differs from its daily counterpart in that the time spans (see Table 15.2) are longer and the formula, like all three included on the chart, uses EMAs rather than the simple moving average. This arrangement facilitates identification of both the direction and the maturity of the primary trend (shown at the bottom), as well as the interrelationship between the short-term and the intermediate trends. The dark plot shows when the long-term KST is above its 26-week EMA and the lighter plot when it is below. Sometimes, KST primary trend signals coincide with the price itself crossing its 65-week EMA, as was the case in December 2007 and August 2009. The chart also shows the approximate trading bands for the short-term KST in bull and bear markets as defined by the long-term KST EMA crossover approach. Note how it rarely moves to the oversold zone during the bullish environments and rarely to the overbought zone when the main trend is down.