Window Oscillators and Overlay Indicators
In fact, this coincidence of buy and sell signals by both oscillators in the
chart example in Figure 8.7 above creates false consensus and may lead the trad-
er into a false sense of confidence. Many oscillators will tend to give the same
buy and sell signals because they are highly correlated, being constructed using
the same underlying price data. This similarity between oscillators is referred to
as multicollinearity. In this case, the RSI is collinear with respect to the CCI. To
mitigate the adverse effects of multicollinearity, it is best to use oscillators that
are derived from data other than price alone. For example, a combination of
stochastics, Chaikin’s money flow index (which incorporates volume), and some
other sentiment or market‐breadth oscillators like the McClellan Oscillator or
Put/Call ratio will yield significantly less collinearity. Many novice traders make
the mistake of putting on a combination of price‐based momentum oscillators
CCI, stochastics, ROC (rate of change), the MACD, and the MOM (momentum)
oscillators all at the same time. This produces an uncanny agreement between
all these oscillator signals in generating buy and sell signals which overwhelms
the novice trader and leads him or her into a false sense of confidence, believ-
ing that this agreement represents a strong signal. In short, collinearity provides
redundant information. In Figure 8.8, we observe the effects of multicollinearity
across the RSI, MACD, and ROC, all of which are displaying similar divergences
and crossover signals.fIgure 8.8 Multicollinearity on the Daily Chart of Apple Inc.
Courtesy of Stockcharts.com