Brian,

I look at the overall bell curve in MCS testing. If I am getting 
upwards of an 80% probability that I will achieve decent average 
returns with a limited average drawdown, for trades within the system, 
then I have found a good system.

Once I find a good system, ONLY THEN do I do the backtesting for 
optimization. More recent data will generally offer better results, in 
predicting the near future - or so others have told me.  I have yet to 
prove that statement on my own.

One thing I should mention is that I no longer spend time backtesting 
for optimization. The reason being that the trader himself (or herself)
is a very large component of the system, and the trader is not 
being "optimized." Therefore, statistically speaking, it makes no 
sense to optimize a system unless the system will be 100% completely 
automated with no human intervention of any kind. This includes 
automated stops and what not.

If you intend on building a "black box" system, then optimize it. 
Otherwise, IMO, your time will be much better spent optimizing 
yourself through learning money management practices that are designed 
to control your emotional response to fear and greed situations. 
Scaled trading methods (see Jesse Livermore) seem to be very good for 
doing this.

This is what works for me. If it does not resonate with you, by all 
means keep looking ;-) I spent countless hours and months backtesting 
for optimizing systems only to ultimately realize it was a waste of MY 
time for any amount of MY discretionary trading.

Hope this helps.

~rhelfer


--- In [email protected], "brian_z321" <[EMAIL PROTECTED]> wrote:
>
> Hello ~rhelfer,
> 
> What is the first one or two things that you look at in 
MCSimulations?
> 
> Can you recommend anyone/anywhere as the best training source for 
MCS 
> applications to trading?
> 
> Brian_z


Reply via email to