[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: SPU program



PureBytes Links

Trading Reference Links

The program was written after trading the same rules in real time for
10 months last year. 
The avg winner was $994.96...avg loser was $1178.08.
I only trade 1 lots in the test(now 2 lots after the split). I don't do
slip because 95%
of the time my entry /exit is violated so I don't do mkt orders. I'm at
a big firm so commissions are
very low.

I really don't know how to use the risk trailing stop or % floor....any
ideas there??

What I've been doing with other systems is optimizing them every 28 days
and starting
fresh P&Ls. Does this make any sense to anyone??


Mitch

> -----Original Message-----
> From:	Scot Billington [SMTP:scot.billington@xxxxxxxxxxxxx]
> Sent:	Wednesday, May 13, 1998 10:30 AM
> To:	MitchT@xxxxxxx
> Subject:	Re: SPU program
> 
> >
> > My problem is ...I don't know how these results compare to others.
> >Take a look :
> >   Net profit $32400
> >   % profitable 62%
> >    192 trades
> >    max drawdown $12850
> >   Return  252%
> >
> >The only stop I use is end of day exit.When I optimize stops
> etc...the
> >results are much higher. 
> >Is using the optimize stops function just kidding myself(
> overfitting)??
> >
> >
> 
> I think further optimization would be curve fitting unless you found
> another virgin set of data on which to use the new parameters.  I also
> bet
> you developed your ideas on this same set on which you tested it.  If
> you
> did, these results are not valid.  You must test on data that was not
> used
> in any way to develop the ideas.  A few other questions I think you
> should
> answer:  How much comm/slippage did you use?  Are you doing around 6-9
> trades per day? (If so, your costs are going to be high.)  What is
> your
> average winner and loser?  How are you deciding position size? 
> 
> You should also make slight changes in your parameters and show
> basically
> the same results to prove robustness.
> 
> sb
>