|
PureBytes Links
Trading Reference Links
|
This is not MY system (I should only be able to construct such!), but I was
passing on some work I had done while experimenting with it. I am too lazy
to search through the postings at the group online, but there was a
commentary last week, something to the effect that the delay was already
hard coded into the formula, therefore it was proper to set the delay at
zero, rather than 1. If, in my ignorance, I have gotten everything screwed
up, please forgive me.
Yes, as you have both pointed out, when you set the delay to 1, the
formula appears no better than every other system I have used to lose money!
Please straighten me out on this matter of the delay already being "built
in" to the formula, and forgive me if I have confused issues with my
ignorance of code. In the meantime, I will continue on the quest for the
Grail :>))
Nate
At 04:38 PM 6/24/02 +0000, you wrote:
Nathan,
I did not run your system on the tickers you presented, but rather on
a watchlist I have here at work. It made excellent money in the years
I tested it in. However, the problem is that you set the delay for
both buy and sell to 0. If you are using EOD data, you can't buy or
sell until tomorrow because you haven't downloaded today's data yet.
So, if you set the delays to 1, you get much poorer results. Try it.
No holy grail yet.
AV
--- In amibroker@xxxx, "Dr. S. Nathan Berger" <snberger@xxxx> wrote:
> Al- 'sHappenen?
>
> Sorry, Monday morning brain fog.
>
> Reference is to using the TRIX formula posted last week to the
group:
>
> pds=Optimize("pds",12,5,23,3);
> pds2=Optimize("pds2",3,3,15,1);
>
> Buy=Trix(pds) > EMA(Trix(pds),pds2);
> Sell=Trix(pds) < EMA(Trix(pds),pds2);
> Short=Trix(pds) < EMA(Trix(pds),pds2);
> Cover=Trix(pds) > EMA(Trix(pds),pds2);
>
> Filter=1;
> Buy=ExRem(Buy,Sell); Sell=ExRem(Sell,Buy); Short=ExRem(Short,Cover);
> Cover=ExRem(Cover,Short);
>
> AddColumn(Buy,"Buy");
> AddColumn(Sell,"sell");
> AddColumn(Short,"short");
> AddColumn(Cover,"cover");
>
> A remark had been made that this formula seemed workable, but that
the
> selection of stocks to which it was applied made a substantial
difference.
> In search for such stocks, I tested this formula over the stocks
listed in
> the CCT50.tls, (which is used by Steve in his CMO System).
>
> When optimizing individually over each stock in the CCT50.tls, the
results
> consistently come up pds=5 and pds2=3. When these optimized figures
were
> entered into the formula, the backtested results for
> AMTL, BEAS, CEFT, AHIR, CHKP, CMVT, CTX, DYN, KLAC, MERQ,
NVDA,
> SEBL. came back returning a performance of OVER 1500% for each
stock. Of
> these, MERQ is 4800% and NVDA is 4900%
>
> Settings: Init. Eq 10,000 Positions: Long Periodicity: Daily
> Long Trade Buy Price: Open Delay: 0
> Short Trade Buy Price: Open Delay: 0
> Max. Stop Loss, Profit Target, Trailing Stop all are Disabled
> Reporting set to Overall Summary
>
> Hope this is clearer now. Also hope this is reality, and not a
continuation
> of the brain fog. Perhaps a cup of coffee would help...
>
> Nate
>
>
>
>
> At 10:08 AM 6/24/02 -0400, you wrote:
>
> >Hi, Nate. Long time, no hear. I'm not sure your post is about the
CMO5
> >system or the Turtle system. Can you elaborate a little on what
you are
> >talking about? What are you setting to 0? How do you use TRIX to
evaluate
> >preformance of a system? I'm at work right now, so I cannot go
back and
> >review past posts. Thanks.
> >
> >Al Venosa
>
>
>
> ---
> Outgoing mail is certified Virus Free.
> Checked by AVG anti-virus system (http://www.grisoft.com).
> Version: 6.0.372 / Virus Database: 207 - Release Date: 6/20/02
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
---
Incoming mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.372 / Virus Database: 207 - Release Date: 6/20/02
---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.372 / Virus Database: 207 - Release Date: 6/20/02
|