Technical Analysis Explained Page 15
The requirements for an exhaustion bar are as follows:
1. The price opens with a large gap in the direction of the then-prevailing trend.
2. The bar is extremely wide relative to previous bars.
3. The opening price develops in the lower half of the bar in a downtrend and in the upper half in an uptrend.
4. The closing price should be both above the opening and in the top half of the bar in a downtrend, and in the lower half and below the opening in an uptrend.
5. The bar is completed with a gap to the left still in place.
Examples of exhaustion bars for both a top and bottom are featured in Figures 10.12 and 10.13.
FIGURE 10.12 Bullish Exhaustion Bar
FIGURE 10.13 Bearish Exhaustion Bar
They differ from the one-bar island reversal, in that while there is a gap to the left of it, there is no gap between the exhaustion bar and its successor. Examples of one-bar island reversals are shown in Figure 10.14.
FIGURE 10.14
What we are looking for here is an extreme movement in the price that is preceded by an already strong move. The idea is that a bullish bar following a sharp decline opens with a huge gap, but closes above the opening and, ideally, more than halfway up the trading range. This reflects the concept of a reversal in psychology. The large gap and wide trading range also point up the kind of frenzied activity associated with a turn. The opposite set of conditions would be present for a bearish exhaustion bar.
Chart 10.10, featuring the 2012 price action of Ruby Tuesday, shows a classic exhaustion bar with the price gapping down sharply following a waterfall decline. Then it closes well into the upper half of the bar, indicating that the sellers were no longer in control. The huge volume also contributed to the idea that an important change in sentiment had taken place.
CHART 10.10 Ruby Tuesday Exhaustion Bar
You will find that gaps almost always develop on the intraday charts at the open due to some overnight change in psychology. This means that exhaustion bars tend to be more prevalent in these very short-term charts.
Pinocchio Bars
Exhaustion also shows itself in forms that are different from those we have so far considered. I call these “Pinocchio bars” because they temporarily give us a false sense of what is really going on. They are bars in which the bulk of the trading takes place beyond a support or resistance zone but the open and close do not. Often, this means the false move takes the price beyond a previous trading range, thereby giving a false impression of a breakout. The character Pinocchio tells us when he is lying because his nose gets bigger. In the case of our Pinocchio bar, it is the isolated part of the bar above the open and close (or below them for a bullish bar) that is the big nose that by the end of the bar signals a probable false move. Figures 10.15 and 10.16 offer two examples of false upside Pinocchio breaks. Figure 10.16 indicates that when a false break develops above a down trendline, this is indicative of exhaustion since the price cannot hold above the strong resistance reflected by the trendline.
FIGURE 10.15 Pinocchio Bars
FIGURE 10.16 Pinocchio Bards and Down Trendlines
Marketplace examples of Pinocchio bars are featured in Charts 10.11 and 10.12. The first shows a break above the trading range, which was nullified by the time the bar closed.
CHART 10.11 S&P Composite Bearish Pinocchio Bar
CHART 10.12 S&P Composite Bullish Pinocchio Bar
The second shows a false break to the downside. As so often happens following whipsaws, the price moved in the opposite direction from that indicated by the break. The outside bar in the left part of Chart 10.12 offers another form of Pinocchio bar.
One important fact about an exhaustion move is that the extremity of the move often proves to be an important support or resistance point. In that respect, it is often a good idea to place a stop loss a little bit beyond the extremity of the Pinocchio bar, provided, of course, that still results in a reasonable risk reward.
Summary
1. One- and two-bar reversals reflect exhaustion and usually signal a reversal in trend.
2. To be effective, these reversals must be preceded by a worthwhile move.
3. Their trend reversal significance is only of short-term duration. What constitutes as “short-term” will depend on the time span of the bar or bars in question.
4. Daily or weekly reversal patterns will be far more significant than those appearing on intraday chart.
5. Reversal bars that contain more of the required characteristics normally provide stronger signals than those that only have a few.
11 MOVING AVERAGES
It is evident that trends in prices for any freely traded entity can be very volatile, almost haphazard at times. One technique for dealing with this phenomenon is the moving average (MA). An MA attempts to tone down the fluctuations of stock prices into a smoothed trend so that distortions are reduced to a minimum. The three principal types of MAs used in technical analysis are simple, weighted, and exponential. When the terms “moving average or MAs” are used in this book, we are referring to the simple type. Exponential MAs (EMAs) and weighted MAs (WMAs) will always be specifically referenced. The construction and use of these averages are different; therefore, each type will be dealt with in turn. In the meantime, it’s important to remember that MAs, like trendlines, should be considered as dynamic levels of support and resistance. They are dynamic because, unlike specific levels, which by definition remain constant, MAs keep changing their values. If a specific MA has not worked well in the past on a particular security, there are few grounds for suspecting that it will in the future and vice versa.
Major Technical Principle Moving averages should be thought of as a dynamic level of support and resistance.
Simple MA
A simple MA (SMA) is, by far, the most widely used. It is constructed by totaling a set of data and dividing the sum by the number of observations. The resulting number is known as the average or mean average. In order to get the average to “move,” a new item of data is added and the first subtracted. The new total is then divided by the number of observations, and the process is repeated.
For example, the calculation of a 10-week MA would follow the method shown in Table 11.1.
TABLE 11.1 Simple MA Calculation
On March 12, the total of the 10 weeks ending on that date was 966, and 966 divided by 10 results in an average of 96.6. On March 19, the number 90 is added and the observation of 101 on January 8 is deleted. The new total of 955 is then divided by 10. The calculation of a 13-week MA would require totaling 13 weeks of data and dividing by 13. This process is then repeated in order to get the average to “move.” Generally speaking, a rising MA indicates a rising trend (market strength), and a declining one denotes weakness.
A comparison of the price index with its 13-week MA (Chart 11.1) shows that the average changes direction well after the peak or trough in the price and is, therefore, “late” in changing direction. This is because the MA is plotted on the thirteenth week, whereas the average price of 13 weeks of observations actually occurs halfway through the week time span, i.e., in the seventh week.
CHART 11.1 NASDAQ 100 2011–2012 Centering an MA
Major Technical Principle Changes in the price trend are identified by the price crossing the MA, not by a reversal in the direction of the MA.
If it is to reflect the underlying trend correctly, the latest MA should be centered, i.e., plotted on the seventh week, as shown in Chart 11.1 for the NASDAQ 100.
The chart in the lower panel shows an MA that has been centered. The good news is that it turns fairly closely to the turning point in the price. The bad news is that it is necessary to wait 6 weeks before it’s possible to ascertain whether the average has changed direction. See how the centered MA plot ends 6 weeks prior to the end of the chart.
A time delay, though it is an irritant, is not particularly critical when analyzing other time series such as economic data. However, given th
e relatively rapid movement of prices in the financial markets and consequent loss of profit potential, a delay of this nature is totally unacceptable. Technicians have found that, for the purpose of identifying trend reversals, the best results are achieved by plotting the MA on the final week.
Changes in the price trend are identified, not by a reversal in direction of the MA, but by the price itself crossing its MA. A change from a rising to a declining market is signaled when the price moves below its MA. A bullish signal is triggered when the price rallies above the average. Since the use of MAs gives clear-cut buy and sell signals, it helps to eliminate some of the subjectivity associated with the construction and interpretation of trendlines.
More often than not, it pays to take action based on MA crossovers, but only if you can look back and see a fairly consistent relationship between the price and the MA. The degree of accuracy depends substantially on the choice of MA, as discussed later. First we need to examine some of the characteristics of MAs in greater detail.
Moving Average Characteristics
1. An MA is a smoothed version of a trend, and the average itself is an area of dynamic support and resistance. In a rising market, price reactions are often reversed as they find support in the area of the MA. If the rest of the evidence agrees, it’s not a bad idea to wait for the price to reach its MA prior to making a purchase. After all, if the MA represents support, you can place a stop below support, i.e., the MA. Similarly, a rally in a declining market often meets resistance at an MA and turns down. The more times an MA has been touched, i.e., acts as a support or resistance area, the greater the significance when it is violated.
2. A carefully chosen MA should reflect the underlying trend; its violation, therefore, warns that a change in trend may already have taken place. If the MA is flat or has already changed direction, its violation is fairly conclusive proof that the previous trend has reversed.
3. If the violation occurs while the MA is still proceeding in the direction of the prevailing trend, this development should be treated as a preliminary warning that a trend reversal has taken place. Confirmation should await a flattening or a change in direction in the MA it-self, or should be sought from alternative technical sources.
4. Generally speaking, the longer the time span covered by an MA, the greater is the significance of a crossover signal. For instance, the violation of an 18-month MA is substantially more important than a crossover of a 30-day MA.
5. Reversals in the direction of an MA are usually more reliable than a crossover. In instances in which a change in direction occurs close to a market turning point, a very powerful and reliable signal is given. However, in most instances, an average reverses well after a new trend has begun and so is only useful as a confirmation.
In short, think of an average as a type of “moving trendline,” which, like and actual trendline, obtains its significance from its length (time span), the number of times it has been touched or approached, and its angle of ascent or descent.
What Is a Valid Crossover?
A crossover is any penetration of an MA. However, close observation of any chart featuring an MA will usually reveal a number of whipsaw, or false, signals. How can we tell which ones are going to be valid? Unfortunately, there is no way of knowing for certain. Indeed, many whipsaws cannot be avoided and should be regarded as a fact of life. However, it is possible to avoid some of these close calls by using filtering techniques. The type of filter to be used depends on the time span in question, and is very much a matter of individual experimentation. For example, we may decide to take action on MA crossovers for which a 3 percent penetration takes place and to ignore all others. Violations of a 40-week MA might result in an average price move of 15 to 20 percent. In this instance, a 3 percent penetration would be a reasonable filter. On the other hand, since 3 percent would probably encompass the whole move signaled by a 10-hour MA crossover, this kind of filter would be of no use whatsoever.
Some analysts, recognizing that one-period whipsaws are quite common, require an MA crossover to hold for at least two periods. In the case of daily data, this approach would mean waiting for the second or third day before concluding that the average had been violated. A more sensible method is to use a combination of the period and percentage penetration for deciding whether a crossover is valid.
A useful tip is to wait for an MA crossover to take place at the same time a trendline is violated or a price pattern completed. Such signals strongly reinforce the trendline or price pattern signal and, therefore, need less in the form of a filter requirement. Two examples are shown in Chart 11.2 for the iShares MSCI Brazil ETF.
CHART 11.2 iShares Brazil 2011–2012 Joint Trendline/MA Violations
Major Technical Principle If an MA crossover takes place at the same time a trendline is violated or a price pattern is completed, these signals strongly reinforce each other and, therefore, need less in the form of a filter requirement.
Sometimes, it’s possible to see an MA crossover accompanied with exceptionally heavy volume. In such circumstances, you could lower your standards of what represented a decisive breakout since the expanding volume would emphasize enthusiasm by the buyers or fear by the sellers, depending on the direction of the break. We see this in Chart 11.3 for an early August 2012 downside break in the iShares MSCI Emerging Markets ETF.
CHART 11.3 MSCI Emerging Markets ETF 2010–2012 MA Crossovers and Volume Characteristics
The cross was accompanied by expanding volume and a trendline break. Two other breaks also developed on expanding activity, one of which was an upside violation, which was also associated with a trendline break.
Chart 11.4 features the Eurotop Index together with a 40-week MA and two bands that have been plotted 3 percent above and below the average itself.
CHART 11.4 Eurotop 1993–2001 40-Week MA and 3 Percent Bands
Buy signals are generated when the price crosses above the upper line and sell signals when it crosses below the lower one. This has the effect of filtering out some of the whipsaws, yet does not give up too much in terms of timing.
MAs are usually constructed from closing data. These are more reliable than intrasession prices because they reflect positions that investors are willing to carry overnight or, in the case of weekly charts, over the weekend. Moreover, bar charts are subject to intraday rumors and other random noise. It’s certainly true that the bar charts are penetrated intraday and, therefore, offer more timely signals. However, there are also a far greater number of whipsaws or misleading signals. That’s because the intraday trading can be subject to manipulation or distorted by kneejerk emotional reactions to news events. For this reason, it is usually best to wait for the closing price to penetrate the average before concluding that a crossover has taken place.
One exception to using close-only charts would be if the MA was touched or approached by the highs or lows of the bars on numerous occasions. In this instance, it would mean that the MA represented an unusually strong resistance or support level. Consequently, its penetration would have greater significance.
During a trading range, MA crossovers have a strong tendency to be counterproductive. Two examples are shown in Chart 11.5 for Asian Paints, an Indian stock. In these situations, it is usually best to use the outer ends of the trading range for the signal rather than the MA.
CHART 11.5 Asian Paints MA 2002–2005 Crossovers and Trading Ranges
Obviously, no one rings a bell to say the price has entered a trading range, but after a couple of whipsaw signals, it becomes apparent. That is the time when a well-constructed trendline should be substituted for an MA crossover.
Choice of Time Span
MAs can be constructed for any time period, whether a few days, several weeks, many months, or even years. Optimal selection of length is very important. For example, if it is assumed that a complete bull and bear cycle lasts for 4 years, an MA constructed over a time span longer than 48 months will not reflect the cycle at all. This is becau
se it smoothes out all the fluctuations that take place during the period and will appear more or less as a straight line crossing through the middle of the data, unless there is a particularly sharp linear trend. On the other hand, a 5-day MA will catch every minor move in the stock cycle and will be useless for the purpose of identifying the actual top and bottom of the overall cycle. Even if the 48-month average were shortened to 24 months, using the crossover signals would still cause the 24-month average to give an agonizingly slow confirmation of a change in trend. The 4-week average would be so sensitive that it would continually give misleading or whipsaw signals. Only an MA that can catch the movement of the actual cycle will provide the optimum trade-off between lateness and oversensitivity, such as the 10-month MA in Figure 11.1.
FIGURE 11.1 A Short- Versus a Longer-Term MA
The choice of MA depends on the type of market trend that is to be identified, i.e., short, intermediate, or primary. Because different markets have different characteristics and the same markets go through different cyclic phenomena, there is no such thing as a “perfect” MA. In recent years, extensive computer research has been done on the optimum MA time span. The conclusion from all sources is that there is no one perfect time span.
What may work extremely well in one market over one specific period of time is unlikely to be duplicated in the future. When we talk about choice of time span, we are really trying to identify an MA that will work most of the time with a specific time frame, i.e., short, intermediate, or long.