Technical Analysis Explained Read online

  The best buying opportunities seem to occur either when the long-term index is in the terminal phase of a decline, or when it is in an uptrend but has not yet reached an overextended position.

  Quite often, the long-term series stabilizes but does not reverse direction, thereby leaving the observer in doubt as to its true intention. Vital clues can often be gleaned from the action of the short-term and intermediate series in conjunction with the price action itself. For example, it is unusual for the intermediate series to remain above or below equilibrium for an extended period—say, more than 9 months to a year. Consequently, when it falls below zero after a lengthy period of being above it, this argues for a bearish resolution to a flat long-term series and vice versa. A sell signal of this nature was triggered in early 2008 after the intermediate KST had been above zero for over 4 years, although you can’t see all of that from this chart. Note also the mega overbought reading in the short-term series that developed in 2009. It was actually the highest reading since the inception of this ETF in the year 2000. The KST can be plotted for free at www.pring.com in this market cycle format for any Yahoo! symbol, whether U.S. or international.

  The KST and Relative Action

  The KST can also be adapted to relative strength lines and is especially useful for long-term (primary trend) analysis when applied to industry groups or individual stocks. This is because sector rotation develops around the business cycle as different groups are coming in and out of fashion. As a result, linear uptrends and downtrends are far less likely to develop than with absolute price data. For a fuller discussion on these matters and KST applications, see Chapters 19 and 22.

  Are There KST Substitutes?

  The KST is not the only answer to our smoothed momentum problems, since it is also possible to substitute the moving-average convergence divergence (MACD) using the default parameters, or the stochastic using a 24/15/10 combination, in the likely event that your charting software does not carry the KST or the ability to replicate it. Those same parameters for the stochastic appear to work for all three time frames. Chart 15.12 compares the three indicators using monthly data. Note that in most cases, the KST coincides with or leads turning points in the stochastic. The dashed arrow indicates when the stochastic led. Alternatively, the more sensitive MACD often gives false impressions of trend reversals that either turn out to be false or are reversed before a crossover of the signal line takes place. For these reasons, I prefer the KST for all trends, not just long-term ones. It may well be possible to come up with superior parameters for both indicators, and the reader is certainly encouraged to make an attempt to do so.

  CHART 15.12 S&P Aluminum Index, 1992–2012 Comparing Long-Term Momentum Indicators

  The Special K

  An alternative method of charting the KST comes when all three of them are combined into one indicator, which I call the “Special K.” This gives us true summed cyclicality where the short-, intermediate-, and long-term trends are combined into one super one. The calculation is made by adding the daily KST formula to that of the intermediate and long-term series based on daily data. Thus, say the 12-month time span used in the long-term KST calculation is replaced by 265 days as an approximation for the trading days in a calendar year. Table 15.3 contains the formula, and Figure 15.1 features the concept in a visual format. It is really a practical adaptation of the dashed line (short-term trend) in Figure 15.1. It also contrasts the slow, deliberate path of the long-term KST with the more jagged trajectory of the Special K. In an ideal world, the Special K ought to peak and trough more or less simultaneously with the price at bull and bear market turning points. In most situations, that actually happens. When it does, the trick is being able to identify these turning points as quickly as possible. The usual caveat with premature momentum turning points developing in linear uptrends or downtrends still applies.

  Table 15.3 Special K Formula

  FIGURE 15.1 The Special K versus the Long-Term Trend

  The prime function of the Special K, then, is to identify primary trend turning points. Since this indicator also includes short-term data in its calculation, a subsidiary benefit lies in identifying smaller trends and putting that in context with the direction and maturity of the primary trend.

  Using the Special K to Identify Long-Term Price Movements

  The following are some of the Special K characteristics:

  1. The Special K is a curve that reflects the dominant long-term KST, but is not as smooth, since it also contains data that reflect short-term and intermediate price movements.

  2. Primary trend peaks and troughs in the price itself often coincide simultaneously with those of the Special K. Where linear uptrends or downtrends are present, the Special K (SPK) leads such turning points and sets up a divergence. The arrows on Chart 15.13 show some of these turning points for the S&P Composite. Note that in 1998 the SPK peaked prematurely due to the presence of a secular uptrend in U.S. equities. There are even more examples in Chart 15.14 of the CRB Composite. Of the 12 peaks and troughs between 1983 and 2012, only one did not develop simultaneously with the SPK.

  3. This indicator lends itself to trendline construction. Usually, when lines greater than 9 months in duration are penetrated, there is a high probability that the primary trend has reversed. This is demonstrated in Chart 15.15 for the S&P Composite, where we see several SPK trendline penetrations confirmed by a similar action by the price. Occasionally, the trend is so steep that it is impossible to construct a meaningful trendline. In such cases, joint moving-average crossovers between the price and SPK that develop within a short time of each other usually serve as timely signals. A joint MA crossover takes place at A along with some trendline violations, but the rally off the 2009 bottom would have been signaled with the MA cross approach. The MA for the price has a 200-day span, and the one for the SPK has a 100-day span smoothed by an additional 100-day MA.

  CHART 15.13 S&P Composite versus the Special K, 1995–2011

  CHART 15.14 CRB Composite versus the Special K, 1983–2012

  CHART 15.15 S&P Composite versus the Special K, 1995–2011

  Chart 15.16 also shows some joint MA breaks at points A, B, and C, as well as numerous trendline combinations. Note that occasionally the SPK will trace out a small trading range, and when this has been completed, reversal signals are triggered. The 2007 and 2010 lows offer two examples of this phenomenon.

  CHART 15.16 CRB Composite versus the Special K, 1999–2012 Trendline Interpretation

  CHART 15.17 BSE FMG Index, 2001–2009 Special K Interpretation

  4. Peak-and-trough reversals often show up at primary trend turning points. We see three instances in Chart 15.17 for the Bombay Stock Exchange FMGC Index at points X, Y, and Z. Not all turning points are signaled in this way, just as it’s not always possible to construct a meaningful trendline. However, if we see a trend break and an obvious peak/trough reversal, the odds of a primary trend reversal are greatly enhanced.

  Using the Special K to Identify Short-Term Price Movements

  If you compare movements in the daily KST in Chart 15.18 with those of the Special K, you will see that they are very close indeed, which means that we really do get the summed cyclicality of that dashed short-term trend curve in our original market cycle diagram (Figure 1.1).

  CHART 15.18 Composite, 2004–2008 the Special K versus the Daily Short-Term KST

  That means we can use its gyrations to help identify short-term reversals in the Special K. For example, if the KST is overbought and reversing, as in October 2007, it is more likely to result in an imminent reversal in the trajectory of the Special K and, therefore, the price. The oversold reading that preceded it in August 2007 also worked quite well. However, during May 2007 we get another KST overbought reversal. The Special K also reverses, but there is no immediate decline. The reason? It’s because this sell signal was a contratrend one, as the prevailing primary trend was bullish.

  Chart 15.19, featuring the CRB Composite, also disp
lays the daily KST in the bottom window.

  CHART 15.19 CRB Composite, 2000–2003 Using the KST Special K Relationship for Interpretation

  We see an overbought KST with a reversing Special K in May 2000, and this is followed by a 3-month correction. However, there is more to it than that. Most of the time when the KST reverses direction, the Special K does as well. However, when the KST reverses to the upside, for example, and there is very little or no Special K response, the chances are that bearish intermediate and long-term forces are dominating, thereby putting substantial downward pressure on the Special K. Under such circumstances, the Special K is most likely going to move lower once the daily KST rally is over, thereby confirming that the main trend remains bearish. Several examples of this phenomenon are seen in the chart, the most glaring of which developed in the August/September 2001 period when the KST experienced a gentle rally but the Special K continued in its decline with no sign of strength whatsoever. This weakness was not apparent by observing the action of the daily KST, but by comparing it to the weak action of the Special K, it was possible to appreciate the downside pressure being applied by the intermediate and long-term cycles. Another small discrepancy developed in July 2001, where the Special K was hardly able to rally at all. The April/May 2001 period also shows a strong KST rally but a very weak Special K advance.

  Bullish divergences develop when the KST declines but the dominant longer-term cycles used in the Special K calculation propel it upward. A good example developed in November 2002 and January 2003, as indicated by the two solid arrows. The idea of rising peaks and troughs for the Special K is especially important, because it indicates strength in the dominant intermediate and primary trend cycles. For example, look at the two dashed arrows. The one for the KST shows a lower low in November 2002 and that for the Special K shows a higher trough, if you can call it that, at B.

  Finally, there is another way in which the near-term movements can help in deciding whether a specific short-term KST buy or sell signal is going to work or not. Note the horizontal dashed line marking the short-term low in the Special K in May 2001. When the indicator violates this level, it signals that a new low in the price itself is likely. In the case of the May 2001 example, the Special K took out its low just about 2 weeks before the CRB itself did.

  The reverse situation developed in November 2002, where the Special K moved to a new high. In this case, there was no lead by the momentum indicator, as the price broke out more or less simultaneously with it.

  Major Technical Principle If the Special K registers a new high or low for the move, the price usually follows.

  How the Special K Generates Short-Term Buy and Sell Signals

  One of the most important things for short-term traders to grasp is the fact that

  Major Technical Principle Trades executed in the direction of the main trend are much more likely to be successful than those generated in a countercyclical way.

  One starting point that helps us arrive at an objective way of determining the direction of the primary trend is to use the SPK. Obviously, we can easily tell with the benefit of hindsight where the actual SPK peaks and troughs formed, but in real time, we do not have this luxury. One solution is to determine its position vis-à-vis its 100-day MA smoothed with a 100-day MA. Positive readings would indicate a primary bull market and vice versa.

  Chart 15.20 shows such a system for the Dow Jones UBS Commodity Index. The shaded areas represent bear markets as defined by the SPK/MA relationships. No signals are generated during this type of environment. The dark highlights indicate when the SPK crosses above its 10-day MA and when the primary bull/bear model is bullish. This approach is far from perfect because there are points when the model is bullish but the price has already reversed and so forth. Rather than blindly using this approach as a mechanical system, I think it better to use pro-trend signals as an alert and to then use other indicators as a filter in a weight-of-the-evidence approach.

  CHART 15.20 Dow Jones UBS Commodity Index, 2009–2012 Generating Pro-Trend Short-Term Buys and Sell Signals

  Self-generated charts and templates for these SPK systems are included in an add-in package for MetaStock, which is available at www.pring.com

  Special K Drawbacks and Benefits

  Like all momentum indicators, the SPK does come with drawbacks. The most noticeable derives from the fact that the indicator’s construction assumes that the price series in question is revolving around the typical business cycle. Consequently, it reverses prematurely during linear trends and will lag when the cycle is unusually brief. That is, of course, a drawback with any long-term momentum indicator.

  However, in the few years I have been working with it, I have grown more and more impressed with its ability to identify numerous primary trend turning points where other indicators have failed.

  The Directional Movement System

  The objective of the directional movement system, designed by Welles Wilder, is to determine whether a market is likely to experience a trending or trading range environment. The distinction is important because a trending market will be better signaled by the adoption of trend-following indicators, such as moving averages, whereas a trading range environment is more suitable for oscillators. In practice, I am not impressed with the ability of the directional movement system to accomplish this objective, other than to identify a change in trend. On the other hand, there are, I find, several other ways in which this indicator can be usefully applied.

  The Calculation

  The calculation of the directional movement system is quite involved, and time does not permit a full discussion here. For that readers are referred to Wilder’s New Concepts in Technical Trading Systems (Trend Research, 1978), to my own Definitive Guide to Momentum Indicators book and CD-ROM tutorial (Marketplace Books, 2009), or to my online audio-visual technical analysis course at Pring.com.

  To simplify matters, the directional movement indicator is plotted by calculating the maximum range that the price has moved, either during the period under consideration (day, week, 10-minute bar, etc.) or from the previous period’s close to the extreme point reached during the period. In effect, the system tries to measure directional movement. Since there are two directions in which prices can move, there are two directional movement indicators. They are called +DI and –DI. The resultant series is unduly volatile, so each is calculated as an average over a specific time period and then plotted. Normally, these series are overlaid in the same chart panel, as shown in Chart 15.21 for the euro using the standard, or default, time span of 14 periods.

  CHART 15.21 Euro, 2011–2012 Featuring a +DI and –DI

  There is one other important indicator incorporated in this system, and that is the average directional movement (ADX). The ADX is simply an average of the + and –DIs over a specific period. In effect, it subtracts the days of negative directional movement from the positive ones. However, when the –DI is greater than the +DI, the negative sign is ignored. This means that the ADX only tells us whether the security in question is experiencing directional movement or not. Again, the normal default time span is 14 days.

  The ADX is calculated in such a way that the plot is always contained within the scale of 0 to 100. High readings indicate that the security is in a trending mode, i.e., it has a lot of directional movement, and low readings indicate a lack of directional movement and are more indicative of trading range markets.

  Major Technical Principle The ADX tells us nothing about the direction in which a price is moving, only its trending or nontrending characteristics.

  The Two DIs

  In Chart 15.21, buy alerts are signaled when the +DI crosses above the –DI (solid arrows) and vice versa (dashed arrows). In this example, there are several occasions when it is possible to confirm such crossovers with a trendline violation in the price. Moving-average crossovers or price patterns could just as easily be substituted. These DI crossovers are fairly accurate and not subject to whipsaws.

&nb
sp; Unfortunately, things do not always work out as well, which you can see from the whipsaw signals in the five ellipses in the bottom window of Chart 15.22. That’s one reason why it is important to make sure that these crossovers are confirmed by the price. Another way around this is to smooth the two DIs as I have done in the middle panel of Chart 15.22 using a 10-day MA. This certainly eliminates a lot of the whipsaws, as you can see from the ellipses, but there is a trade-off in that the signals are occasionally delayed because this approach is less sensitive. In this instance, I have placed arrows at some of the key points where the smoothed series cross each other. Again, it’s important to remember that these are momentum signals and should be confirmed by the price. In addition, this approach, like most others, should be used where you decide to fight the battle. That means that if the price has already moved a long way by the time the signal has been triggered, it is probably best to ignore it for new positions.

  CHART 15.22 Gold Trust ETF, 2011–2012 Comparing “Raw” with Smoothed DIs

  The ADX

  A high ADX reading does not tell us that that the market is overbought and about to go down. Instead, it measures the intensity of the move from a directional point of view. Consequently, when it reaches a high reading and starts to reverse, a warning is given that the prevailing trend has probably run its course. From here on in we should expect it to change. This is different from a reversal in trend, since a change in trend could also be from up to sideways or down to sideways. In Chart 15.23, a 14-day ADX has been plotted against the Gold Trust ETF, the GLD. Note how reversals from high readings signal turning points at both tops and bottoms. Since the upward and downward trajectories are fairly deliberate, negative 10-day MA crossovers at high ADX readings act as good confirmation that a peak has been seen. The indicator in the bottom panel is simply the differential between a 10-day MA of a 14-day +DI and a –DI, as shown in the center panel of Chart 15.22. By comparing the position of the ADX to the differential, it is fairly obvious whether a high-end reversal is coming from an overbought or oversold level. When the ADX reverses in such a manner, it’s then a good idea to obtain some kind of price confirmation. Of the three instances in this chart, only the first in September 2011 was confirmed.