Divergence Analysis
- One data series is down while the other is up
- One data series is down while the other is sideways
The following three combinations are regarded as moving in the same general
direction:- Both data series are moving up
- Both data series are moving sideways
- Both data series are moving down
9.1.3 Determining Direction in Divergence analysis
When trying to ascertain whether a data series is moving up or down, we examine the:■ (^) Sequence of peaks and troughs (or swing highs and swing lows, respectively)
to identify the direction of the data series over a specified number of cycles or
oscillations
■ (^) Slope (or angle) between two data points to identify the direction of the data
series over a specified range or duration of data (especially in the absence of
visible peaks and troughs)
It must be noted that in order to determine direction by comparing peaks
and troughs, the formation of the subsequent peak or trough must first be firmly
established. A peak or trough is usually defined and established via the application
of some kind of filter. As seen in an earlier chapter, the establishment of peaks or
troughs is strictly based on price filtering. Time‐based filters are not applicable, as
a certain amount of price excursion is required to establish a new peak or trough.
It is interesting to note that some practitioners also resort to using regression
lines in order to establish the direction of the data series, in the hope of achieving
greater consistency when analyzing divergence. But it will be shown that such an
approach will only lend itself to standard and not reverse divergence analysis.
Before proceeding any further, it should be stressed that establishing direction
in divergence analysis differs slightly from that of conventional trend analysis,
having basically four ways to establish direction. Generally speaking, the deter-
mination of direction depends exclusively on the definition being employed; for
example, an uptrend may be defined as a series of higher peaks and troughs, while
a downtrend may be depicted as successively lower peaks and troughs. Notice
that both peaks and troughs must be moving in the same direction in order to
be deemed a trend. However, in divergence analysis, as will be described shortly,
it suffices that the direction of the current larger trend may also be determined
solely by comparing adjacent peaks or troughs. Note that this unique definition of
direction is only applied when attempting to determine the direction of the current
larger trend across a single oscillation at the lower wave degree and is never used
to determine the direction of the prior or subsequent larger trend, both of which
are defined via the conventional approach to trend analysis. Therefore directional
discrepancy between the data series at the current larger trend may be based on a
very specific definition, as opposed to the conventional definition of trend analysis.