On Composite Sentiment
by James P. Martin
Despite all the powerful technical tools available, even the savviest technicians are susceptible to
emotion and on occasion succumb to psychological influences that prevent them from taking appropriate
action, thus losing out on opportunity. Nevertheless, James Martin writes, how the market reacts is a
fascinating subject. He continues his quest for more quantifiable measures for sentiment indicators.
In my first article in the June 1990 STOCKS & COMMODITIES, I showed a refinement of traditional call/put
ratios that greatly enhanced their reliability. (See "Using option ratios," elsewhere.) My second in
February 1991 delved further into the qualitative interpretation of the ratios and how they should be
combined with sentiment surveys and the media. Numerous readers called me, hoping to find some
magic mechanical signal that would tell them just when the market was about to turn. I explained that in
my sentiment studies I have experimented with simple moving averages, exponential moving averages,
envelopes, Bollinger Bands, rates of change and other methods I would be embarrassed to admit that I
tried in public. While my studies have produced some general guidelines, I have yet to discover anything
purely mechanical when trading off sentiment indicators.
The heart of the problem lies in determining when traders have truly put their money where their mouths
are. Is an opinion survey reading at, say, one standard deviation coupled with a call/put ratio at plus two
standard deviations a significant contrary opinion signal? Or is it merely a sign that a handful of savvy
professionals have made sizable bets in the options market, thereby skewing the options ratios?
Conversely, if sentiment surveys move sharply but option activity remains subdued, is it a sign that traders are already committed and are now merely "talking their own book"? Or will options traders be
drawn in further before the market turns? This dilemma has led many traders to dismiss sentiment
indicators as being too subjective to be considered valuable timing tools.