Forum Archive Index - January 2004

[robin benson <rob@hammerheadmedia.co.uk>]

Interesting topic Phaedrus.

One simple observation/general comment: recognition of a trendline and 
signaling breaks, prompting some action (buy/sell), would surely be a 
complete waste of time on the basis that if it works, many people would 
have access to the same software and the edge is lost. Many successful 
traders are continually evolving and changing their systems for this 
very reason. So the non-mechanical component of the system is 
absolutely critical.

Robin

On 15 Jan 2004, at 19:37, Phaedrus wrote:

> T100,
>      I have not had any real success applying totally mechanical 
> systems to the local market. I have had limited success with Aus 
> stocks in that my systems only work reliably with the most heavily 
> traded stocks there. My best results have been obtained with ETF's in 
> the US. I suggest you concentrate initially on the QQQ and/or SPY.
>  Some stocks do not seem to be suitable for this type of trading. For 
> example, I have spent (wasted) an inordinate amount of time trying to 
> create a reasonably reliable system for Telecom (TEL/NZ) without a 
> glimmer of success. My theory is that this stock is used as a proxy 
> for the NZ dollar by major traders, and this "overlay" interferes with 
> the basic system. Maybe.
>  You do not have to go too far into backtesting/optimisation routines 
> before you start to run into the limitations of MetaStock and require 
> a more powerful and sophisticated program such as Tradestation. 
> OmniTrader is good for comparing backtesting and forward testing 
> results, though it has nothing like the flexibility of MetaStock.
>  It would take very powerful software indeed to recognise a trendline 
> and signal a break. Some quick easy rough alternatives are to use 
> moving average crossovers, moving average slope, linear regression 
> slope and suchlike.
>  Systems can be optimised for whatever you want - maximum profit, 
> minimum number of trades, best hit-rate, best profit/loss ratio etc 
> etc. One factor that should not be overlooked, in my opinion, is the 
> number of consecutive losing trades. I have had good profitable 
> systems that I personally found unuseable because of their 
> characteristic of periodic runs of losing trades. I can stand 3 or 4 
> consecutive losses, then I begin to lose faith in the system.
>  There are some good books on this subject, they include :-
> Design, Testing, and Optimization of Trading Systems by Robert Pardo.
> Tradings Systems That Work: Building and Evaluating Effective Trading 
> Systems by Thomas Stridsman.
> The Encyclopedia of Trading Strategies by Jeffrey Owen Katz, Donna L. 
> McCormick.
> The Ultimate Trading Guide by John R. Hill,George Pruitt,Lundy Hill
>  Here is an introductory article :-
> http://www.snapdragon.co.uk/resource/articles/nealh.htm
>
>  Hope this helps,
>                   Regards,
>                             Phaedrus.


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