Forum Archive Index - November 2001

[Phaedrus <Phaedrus@techemail.com>]

"If TA is less effective on lightly traded shares, what is the ideal threshold?"
Alec, there is no cutoff point as such. The more heavily a stock is traded, the 
better TA works. On average I am more successful in Australia than NZ, and more 
successful in the USA than Australia. Volume alone is not the key - a million 
one cent shares do not equate with a million $100 shares. Market capitalisation 
is not the key - similar sized companies can have vastly different daily share 
turnover figures. Even the relationship between size/volume/turnover and 
effectiveness of TA is tenuous. The simple fact is some stocks trade well with 
TA and some don't. An example would be Telecom. Ample volume and turnover etc, 
but TEL remains a difficult stock to trade well, technically. At the other end 
of the scale, a stock like VTL which I regard as being too small for serious 
trading, performs quite well with simple trend following systems. So, how do 
you decide? There are two ways that I know and use. (1) The look of the chart. 
Does the stock trend up and down in a nice steady manner, or does it jump 
around a lot, with big unexpected steps up and down? Does the daily chart show 
nice clear candlesticks, or are there lots of four-price-doji? Can you draw 
multiple good trendlines, preferably with multiple confirmations? Does it look 
as though a simple moving average crossover system would be profitable? Do the 
lengths of the zigs and zags suit your preferred trading frequency? Are you 
able to find indicators that give more good signals than bad ones? (2) 
Backtesting. It is possible to backtest any number of systems or indicators on 
any stock, optimising parameters for each system and stock. Any system that 
fails to give good results on backtesting can be guaranteed to lose money. 
Unfortunately good backtest results do not guarantee future success, but they 
do increase the odds substantially. The target here is that any system should 
outperform buying and holding by a significant margin. If it can't, chose 
another stock or another system. Some stock/system combinations can give 
spectacular results, by both backtesting and forward testing. 

  Alan, I totally agree with your views on GPG. The real test of a technical 
analyst is this - are you able to follow a system, even if you think it may 
have given a wrong signal? You must be true to your system - if you start to 
"second guess" it, you are in trouble. Keep to it or change it. The simple 
trendline based system I was demonstrating gave a clear sell signal - I was 
honour bound to "act" on it. Good on you spotting the ascending triangle. Any 
breakout above this would have to be a buy.
       Regards,
                    Phaedrus.

_____________________________________________________________
Are you a Techie? Get Your Free Tech Email Address Now! Visit 
http://www.TechEmail.com


----------------------------------------------------------------------------
To remove yourself from this list, please use the form at
http://www.sharechat.co.nz/chat/forum/