optimization over a portfolio: suggest abandoning TradeStation

[Mark Johnson <janitor@xxxxxxxxxxxx>]

Tom Feldberg <tom@xxxxxxxxxxxxxxxx> writes:
  > Suppose I trade a system with two (or more) optimizable
  > parameters over ten markets using TS.
  >
  > I would like to optimize the parameters for the
  > portfolio rather than for the individual markets. 
  >
  > Also, I would like to be able to chose for which
  > performance measure I am optimizing, e.g., profit,
  > drawdown, profit factor, etc.

I doubt you'll like my reply, but when faced with
exactly this problem in my own trading, I have
solved it (for myself) this way: DON'T USE
TRADESTATION.  I have found that both "System Writer
Plus", by Omega Research, and "Trading Recipes",
by RW Systems, do this job.  Whereas TradeStation
does not.

For the optimization criteria that Tom mentions
(profitfactor, drawdown, profit), I feel that
S.W.P. is superior.  For other optimization criteria
(such as Compound Annual Growth Rate or optimal-f),
I feel that Trading Recipes is the winner.

So here's what I do: I perform all my research
and testing using non-Tradestation software.
Then when it comes time to hand the system over
to a RoboBroker to trade it for me, I port the
code to Tradestation and mail an ELA file
to the broker.

For *me* (notice I'm not saying for *you* or
for anybody else, just me) TradeStation is nothing
but an accepted lingua franca that lets me
communicate my system to a RoboBroker.  For me,
Tradestation is _not_ a research platform or a
backtesting engine.  It's just a reference
implementation of Easy Language which, for some
reason, the RoboBroker industry has standardized
upon.  And if I want to use a RoboBroker, I have
no choice but to employ EasyLanguage to encode
my trading instructions.  I do, so I do.  :-)

   Mark Johnson     Silicon Valley, California     mark@xxxxxxxxxxxx

   "... The world will little note, nor long remember, what is said
    here today..."   -Abraham Lincoln, "The Gettysburg Address"