|
PureBytes Links
Trading Reference Links
|
Jim,
See imbedded answers.
--- In amibroker@xxxxxxxxxxxxxxx, "Jim Ellis <jhellisjr@xxxx>" <jhellisjr@xxxx> wrote:
> Fred (or anyone),
>
> Thanks for posting your system. I've had some time to play with it
> and have a few questions...
>
> You commented earlier in previous posts that you do not trade stocks
> or ETF's...what vehicle do you prefer? Do you trade both long and
> short?
I typically trade mostly MF’s at this point. There are short oriented MF’s which I sometimes trade. My view on ETF’s is that all but a couple of them are too thinly traded to be viable. MF’s take the edge off things like earnings surprises and allow one to grow capital without as much risk i.e. lower drawdowns. My personal belief is that as one gains experience and capital the emphasis shifts from “grow it as fast as you can” to “protect what you’ve got” and as such focus on trading vehicles shifts from futures to options to stocks to mf’s to ?
>
> You chose an oscillator similar to the one Steve Karnish had posted
> earlier. Any comments on the relative merits of the two StoRSI
> oscillators?
>
> Karnish=StoRSIx=100*EMA((RSIa((Price),(P=3))-LLV(RSIa(Price,P),P))/
> (HHV(RSIa(Price,P),P)-LLV(RSIa(Price,P),P)),3);
>
> Tonetti=StoRSIx=100*Sum((RSIa((Price),(P=3))-LLV(RSIa
> (Price,P),P)),3)/Sum((HHV(RSIa(Price,P),P)-LLV(RSIa(Price,P),P)),3);
They’re similar but different, try plotting them both.
>
> It's hard to get my head wrapped around the logic associated with the
> momentum component of your system. It seems like this logic is
> exactly the opposite of the OBOS component. Can you elaborate on
> what market characteristic you are trying to capitalize on with the
> momentum component?
It is the exact opposite. The momentum or trend following part is what you are likely to find in most systems, IMHO it’s the OB/OS component that makes it semi-interesting and that this system employs both.
>
> With 9 parameters to optimize, it seems that this system has a enough
> variables that allows it to curve fit the data over the life of the
> Q's fairly well. Is it practical in your opinion to trade a system
> optimized around 9 variables and expect it to give similar results IN
> THE FUTURE? As you said, and as I found, the parameters to
> be "fragile" with the DD's being hyper fragile.
I never commented on the fragility of the DD’s only the parameters. To me the more fragile the parameters the less robust the system is. The practicality of trading systems with more or less variables is to me not the issue, it’s the robustness of the system.
>
> It seems to make this a robust system where there are several
> parameters in the vicinity of the best that give similar results that
> we've got a ways to go? Did your big brother system reduce the
> number of variables? What strategy did you use to make this concept
> more robust and less sensitive to individual variables?
The big brother did not reduce the number of variables.
>
> You commented in 34239 that it's big brother performed well in OOS
> testing. Can you post only the OOS test results? The big profits
> that are part of the optimization period that are included in the
> overall test period that you posted and make the results hard to
> interpret in the real area of interest which is the OOS period.
I no longer have the parameter settings from the in sample optimization run to generate the out of sample results.
>
> It's interesting that this system performed better with NDX used as
> price rather than QQQ's close. Do you have any idea why?
The results are pretty similar.
>
> Finally, your system is not a true reversal system and the one that
> Herman posted is and produces higher return because it's always in
> the market. What do you think about the pros/cons of the true
> reversal system and why did your original system have an out of the
> market component.
I personally don’t like always in type systems. It had the out of market component because it worked better. If one removed that component one could eliminate half the variables but try your own optimizations with it eliminating the long/short exits and see if you can produce similar results.
>
> Thanks for sharing and the stimulating discussion.
>
> Best Regards,
> Jim Ellis
Yahoo! Groups Sponsor
ADVERTISEMENT
Send BUG REPORTS to bugs@xxxxxxxxxxxxx
Send SUGGESTIONS to suggest@xxxxxxxxxxxxx
-----------------------------------------
Post AmiQuote-related messages ONLY to: amiquote@xxxxxxxxxxxxxxx
(Web page: http://groups.yahoo.com/group/amiquote/messages/)
--------------------------------------------
Check group FAQ at: http://groups.yahoo.com/group/amibroker/files/groupfaq.html
Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.
|