Although I don't agree with all of Phillip Coggan's assertions, it is important to address his critisisms and accept them where they are reasonably fair. According to scientific doctrine critisism adds strength to a theory.
Phillip Coggan asserts the following:
1) the Coppick indicator has provided 2 unreliable signals recently and is therefore unreliable.
TA is used to raise the odds in your favour not as a means of providing a sure signal. Users know better than anyone that TA is often wrong and should not be relied upon without a stop loss of some kind. The Coppick is just an entry signal anyway and is therefore not a complete strategy. In fact TA is often used merely as an entry signal.
2) Coggan agrees that price momentum appers to persist, Coggans critisism is aimed at the more elaborate patterns and systems.
Yes, there are many studies which suggest momentum sytems provide good entry signals, but does this not make relative strength (not RSI) a valid TA indicator?
3)No candlestick patterns (such as three black crows) or technical indicators such as head and shoulders have passed tests of statistical significance
David Schwartz has analysed head and shoulder patterns in an objective way and found statistical significance for data based on LSE share patterns, although certain configurations are more significant than others.
I have never seen the most successful methods such as trends in combination with conforming tops/bottom breaks of trend analysed. I do believe there are objective ways of doing this.
4)Technical analysis is rife with so many subjective intepretations:
I agree with him here and only methods which can be subject to objective analysis should be taken seriously.
5)Elliot waves and Kondratieff cycles are too long term to be proven
This is true of Kondratieff cycles but not of shorter term Elliot waves, only the longer term versions are difficult to prove. I agree however the fractal nature of the markets makes it particularily difficult to provide definitive proofs.
6)As people react to patterns their behaviour can change
I have often suspected this and users of any system should be aware that feedback can potentially invalidate what appears to be a valid system, be it based on FA, TA or otherwise. There is no sure long term method in the stockmarket.
7)Financial markets cannot be predicted by one off events such as September 11th
True, therefore one should never be overexposed.
8)We often see patterns which are not there....as stated in the excellent book by Nasim Taleb
That is why we should try to be as objective and critical as possible, and yes it is an excellent book.
One of Coggan's sources of information 'Practical Speculaion by Niederhoffer and Kenner' may not be what TA enthusiasts wish to read but sometimes you can learn more from a sceptical viewpoint than someone preaching to the converted. It doesn't mean the sceptic is right in all respects but they may be able to highlight points which those fully committed to the TA cause have previously overlooked.