The FinancialContent Network     SocialPicks Community   |   MarketMinute Monitor   |   MarketMinute Market Updates   |   MarketMinute Stock News
SocialPicks
   Sign Up   |   Log In   |   What is SocialPicks?     

The Validity of Head and Shoulders Patterns

 Jul 16, 2009 04:14 PM UTC
Symbol Sentiment Start Return Closed
TA n/a

Graphic_arrow1 Via Condor Options:  

At Friday’s close, technicians were generally pretty bearish on equity indexes, based on the head and shoulders pattern identified below. (Click images to enlarge.)


spx-7-10-09


Of course, this particular setup hasn’t panned out well so far:


spx-7-15-09


In the jargon, yesterday’s rally decisively violated the neckline and pushed up to attempt a retest of the right shoulder. I am unsure whether there are any vaunted technical analysis conventions allowing for mutant patterns featuring two right shoulders, but yesterday’s move certainly sent some analysts scrambling.


Long time readers know that I’m not the world’s greatest fan of technical analysis (TA), at least as it’s practiced by a majority of its proponents. In trading, subjectivity is your enemy, not your friend, and far too many technical analysts rely on interpretive, intuitive, and unquantifiable factors when making decisions. Contrast this with what David Aronson calls evidence-based technical analysis, that is, analysis that can be formalized and tested empirically on historical data by a computer. Subjective TA strikes me as an unverifiable waste of time and money; any particular objective TA indicator might also be useless, but will have the insuperable advantage of being testable on that point.


Head and shoulders patterns like the one above might look like perfect candidates for dismissal as subjective bunk. But apparently in 2000 Andrew Lo and some colleagues developed an approach for automatic pattern recognition of some favorite setups from the world of subjective TA. The literature since then has been mixed on H&S patterns, with criticism from Carol Osler and support from Savin et al. Depending on whom you believe in that literature, this most recent failed pattern might be the exception that proves the rule.






 Graphic_website1 Read the rest of original post »



Add Comment

Be the first to comment on this story and earn 2 points.

Your Comment



IN THE PRESS
Press_forbes Press_washingtonpost Press_wsj Press_npr Press_techcrunch