Weekend Charting and Chat - 21st October 2016, page-9

Hi GJ,

It's a well known adage of the Efficient Market Hypothesis (EMH) that any "edge" will eventually be arbitraged away.

So things may work for the trader for a while, but eventually the "edge" will be reduced to zero. How long is "eventually"? Moot point.

Also, let me be clear, I don't think that knowledge of candlestick patterns is useless as, I believe, you interpreted my post.

I did believe that my post referred only to single candlesticks only. This was the point of Bulkowski's research into Dragonfly dojis. After effects of Dragonfly dojis, he points out, are random, which is at odds with the received wisdom.

I'd also point out that his conclusion that after effects of Dragonfly Dojis are random occurred back as 1999 in his book "Encyclopedia of Chart Patterns". It has been confirmed time and time again by his more recent research.

The extensive testing of single candlesticks by the Massey University has also confirmed the random after effects of such candlesticks.

I also pointed out that, currently, Abandoned Baby (and its inverse "Evening Star") currently have a high reliability. These are, of course, not single candlesticks but groups of candles. (Sometimes, the "Evening Star" pattern is described as a "Bearish Abandoned Baby").

So - to sum up. I do believe that certain groups of candlestick patterns are useful when interpreted in context of volume, trend, and support/resistance.

This is a complex matter where art overtakes science. I'm not sure that computers are yet able to take into consideration all such factors and determine outcomes.

Certainly it is the case, that university researchers haven't been able to investigate such matters scientifically despite their access to high quality computing power. Invariably, university researchers generally isolate single factors and investigate those - not large multiples of factors.

So - it could be some time before the art of the experienced technical analyst can be arbitraged away.

RB.