Analysis Methods of Trend Confirmation

The methods for analyzing confirmation rely on two main tools: indexes and oscillators. Indexes are similar to the breadth line described in the market indicators discussion in Chapter 8, “Measuring Market Strength.” They are cumulative sums of data, usually some variation of volume and price that continuously measure supply and demand over time rather than over a specific period. They do not have an upper or lower bound and are plotted with price charts where they can be compared with price action. The level of the index is irrelevant. What is relevant is the trend of the index relative to the price trend. The only useful indexes are those that begin to change direction before prices, signaling a change in trend. The analyst, thus, compares the prices with the index, looking for divergences between highs and lows in each. Although indexes can also be used with trend line, channel, and occasionally pattern analysis, their most important use is that of divergence analysis in trending markets.

Oscillators, on the other hand, are often bounded and limited to a specified past period. As shown in Figure 18.1, they tend to oscillate within these bounds and demonstrate when volume or prices are relatively high or low. These indicators show the relative changes rather than the absolute changes illustrated in indexes and are also amenable to divergence, trend line, and pattern analysis. Oscillators are used more successfully in trading range markets.

As with many technical indicators, research has not shown that indexes and oscillators are profitable on their own. The student must judge after thorough testing and experience whether to use them as secondary indicators to price analysis. They can be used to generate signals but must always be confirmed by actual and corresponding price action.

Various visual techniques have been developed to determine some meaning from these indexes and oscillators. They are divided into methods that are peculiar to indicators, such as divergences, and those that are just an extension of classical chart pattern analysis, such as trend lines and support and resistance. The principal methods unique to indicators are discussed in the following sections.

1. Overbought/Oversold

Oscillators can be bounded or unbounded. Bounded means that the oscillator swings back and forth within certain bounds or limits. These limits are the extremes to which the oscillator can reach. In most bounded and in some unbounded oscillators, a zone is chosen to represent the range near the extreme bounds. The oscillator might not reach the actual extreme bound, but it might come close and by doing so have the same implications. The zone then is the range that is close enough to the extreme bound to be important. The upper zone is called overbought, and the lower zone is called oversold. When a security has risen far enough that its oscillator reaches the overbought zone, it is said to be overbought, and when the price has fallen far enough that the oscillator reaches the oversold zone, it is said to be oversold.

In a trading range, the overbought and oversold levels are excellent indications of potential reversal levels, especially when the oscillator breaks out from the zone. The zones, however, can be deceiving in a trending market because the oscillator will remain in them during the period of the trend, and, thus, many breakouts from the zone will be false signals. In the following description of oscillators, we point out the conventional zone levels, but as always, the analyst must test for those most appropriate to the time and security being traded.

2. Failure Swings

A failure swing is a specific type of breakout from an overbought or oversold zone first described by Wilder (1978). A stronger version of the breakout, it often is the first sign of a potential reversal after a lengthy trend in which an oscillator has remained within or close to a zone. A negative failure swing is shown in Figure 18.1; it occurs when the oscillator breaks down out of an overbought zone, creates a reversal point, pulls back but fails to reenter the zone, and then breaks back down below the earlier reversal point. A positive failure swing is the opposite of an oversold zone.

3. Divergences

Although Wilder is also credited with discovering divergences, the concept is at least as old as Dow Theory. The basic concept is that to ensure a trend has begun or is still strong, all methods must be confirming the trend. If one index, for example, is breaking out upward but other indexes are not, the indexes are said to be diverging; in other words, they are not acting in concert with each other. A legitimate, strong trend, however, should have all indexes acting in concert. A divergence is, thus, considered to be a sign—especially after a trend has been in existence for a while—that the trend is slowing down and preparing to reverse. We saw breadth divergences in Chapter 8 and how important and reliable they have been. This is the same concept applied to oscillators and price charts.

The basic rule is that when a price reaches a new high, the oscillator should also reach a new high. Of course, this is subject to some interpretation, namely “what high is a new one?” but generally the relative highs should take place at the same time. A high on day 4 and another on day 20 should appear in both the price and the oscillator. If the price high on day 20 is higher than the high on day 4, but the oscillator high on day 20 is not higher than on day 4, the pattern is called a negative divergence. An example of negative divergence is shown in Figure 18.1. A positive divergence occurs at a series of price lows when the price reaches a new low unconfirmed by a new low in the oscillator. In a particularly strong trend, several divergences can occur and, of course, if the oscillator is bounded, it will occur more often because there is little room for the oscillator to keep making new highs or lows. This is why the first negative breadth divergence is often a false signal, but when two occur, back-to-back, the oscillator breakdown signal is more meaningful.

4. Reversals

Brown (1999) describes a variation of divergences known as oscillator reversals in detail. Like divergences, reversals have been used by technicians since oscillators were first used, but they have been popularized recently by Andrew Cardwell. A reversal differs from a divergence in that price leads the change instead of the oscillator. For example, a negative reversal occurs when, on day 20, the oscillator reaches a new high above that of day 4 but the price does not. As illustrated in Figure 18.1, it is sort of a divergence in reverse in that the price is showing weakness when the oscillator is not as opposed to the divergence when the price is still strong but the oscillator is not. Nevertheless, in keeping with the concept of confirmation, the two factors, price and oscillator, are not in sync and are, thus, no longer confirming the trend. A positive reversal is the same as the negative only at low bars. A reversal has the same implications as a divergence— namely, that the trend is beginning to show signs of stress and potential reversal.

5. Trend ID

Brown (1999) also describes what she calls “trend ID.” In a trending market, oscillators will remain in one half of their range for long periods, and breakout signals from the standard overbought and oversold zones are often false. For example, when a price is trending upward, the oscillator can remain at or close to the overbought zone and never reach the oversold zone to give a buy signal during corrections. Brown suggests that the zones should be redefined with the zone parameters raised to include those corrections at a slightly higher level. Her work centers on what is called the “RSI,” a bounded oscillator we look at later, that traditionally has an oversold zone below 30 and an overbought zone above 70. In a strongly upward trending market, the oversold price corrections at which opportune trades might occur are never reached by the oscillator. Thus, she suggests that during the rising trend, the oversold zone be raised to 40 and the overbought zone be raised to 90. Better signals then occur using these new levels as long as the underlying trend remains strong. In a downward trend, the zone levels can also be adjusted downward for the same reason. This adjustment of zones has no effect on divergences or reversals or any of the other chart patterns that might suggest an upcoming trend change. The analyst, however, must test for the best zones to fit the trend.

A variation of trend ID that is commonly used is the standard deviation of the oscillator, similar to the use of standard deviation in Bollinger Bands. The oscillator is calculated and two bands, upper and lower, surround its plot at the level of some multiple, usually one, of the oscillator’s standard deviation. This gives a “moving” overbought and oversold that trend with the oscillator and adjust to changes in the oscillator’s volatility. Signals are generated when the oscillator breaks out of the overbought or oversold zone toward the center just as with the classic zones.

6. Crossovers

Crossovers occur when the oscillator crosses over either a particular level or another oscillator. One level that is often important is the middle value, usually either zero or one, which bisects the range of the oscillator’s travels. Almost by definition, when an oscillator reaches and remains above or below the midpoint in the oscillator range, it is defining the underlying trend. It is, thus, a potential trend indicator. Other oscillators have their raw figures smoothed through a moving average, and crossovers occur when the raw figure crosses over the moving average. These crossovers can either be signals to act or indications of trend change.

7. Classic Patterns

Strangely, oscillators and indicators often make simple patterns, such as triangles and rectangles, and produce support and resistance levels just as price does. They even have trends that can be defined by a classic trend line. These patterns have the same validity as price patterns, even when the oscillator is bounded, and can be additional evidence of trend change or short-term opportunity.

Source: Kirkpatrick II Charles D., Dahlquist Julie R. (2015), Technical Analysis: The Complete Resource for Financial Market Technicians, FT Press; 3rd edition.