Signal clarity

[Doji3333]

One other thing to consider...

Take a look at the TRADES result for any one of the backtests...

Has anyone noticed that a good portion of the LOSSES are caused by weekends. Nearly every single weekend results in a $400 to $1200+ LOSS.

I would have to assume that closing the trades on FRIDAY and opening back up for trading on Monday at 12AM (or whenever) might dramatically improve your trading results.

By a rough estimate (totaling up the trade results "in reverse" for all of the weekends in the following backtest report, the results were +$2037.50 added to the total.

http://share.esignal.com/ContentRoot/Three%20Signal%20Backtest%20Results/13RSI.html

Now, these results are probably inaccurate (as I'm simply eliminating trades that spanned the weekend). But this might be a key to better success.

Brad

[dloomis]

Doji3333

David, Alex or FibGann.. if you could post the most recent code you are using to backtest - I would be willing to try to add my two cents.

Here it is, thx for the comments so far..

[ACM]

Attached is a revised version of FGallfalse that adds function parameters to all the indicators (except the MACD).
Hopefully this should make it a bit easier to cycle through the settings.
Alex

[dloomis]

Re: Reply to post 'Signal clarity'

FYI

I just got a sell signal on the FGallFalse system at 172.75 - 2.25 pts below
the last buy signal.


----- Original Message -----
From: <[email protected]>
To: <[email protected]>
Sent: Tuesday, December 16, 2003 10:57 AM
Subject: Reply to post 'Signal clarity'


> Hello dloomis,
>
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>

[FibbGann]

Dloomis, in regards to your last post. This is what I found wrong with the efs. This is what I am churning on, If we get a buy signal, our indicators could go all the way up to just below our crossover line (sell signal) and come back down without receiving a sell signal.
Still plotting how to fix this in the most efficient way.

Fibbgann

[Doji3333]

My first addition...

Hey Guys...

First off, very interesting... I made a few minor changes and illustrated a few simple things regarding entry triggers..

Changes....

1. I added an "EntrySwitch" that allows users to select..
0 = default "Dloomis" mode
1 = BLM Alternate Entries (basically reversed the logic of the dloomis entries).
2 = BLM Short Term Entries - more of a scalping system

2. When the EntrySwitch is changed to 1 or 2, the code automatically changes the RSI value from 21 to 11 (one of my favorite settings.

3. When the entry switch is changed, the TOTAL limit is reduced to 2 (from 3).

4. Added a variable called "BLMAltOffset" (defaulted to -8). This is used for the type #2 entries. It compares the movement of the CCI and RSI indicator NOW against the value -8 bars ago. This is a method of tracking market trends. We could expand on this a bit by tracking pivot highs and lows in these indicator for potentially better signals.

Now, you may ask "why have som many different entries"??

Answer - exploring the "root" of the system for variances is important to resolve the "best solution".

One thing I noticed with the original system is HUGE DRAWDOWNS - in some cases 10+ pts. So, I'm attempting to find solutions to resolve this and to improve the results of the system.

Comment : EntrySwitch = 2 is a short term system (scalping mode). I'm thinking something like a 1.0 pt profit target may work well here. Entry with 2 contracts, exit 1 contract at a minor profit target and let the rest run. We'll have to see..

David, Alex and FibGann, I'm just offering some "different" version of the same code. I hope I'm not stepping on anyone's toes here.

BTW, I just added an entry switch #3 - checking current CCI and RSI agains highest high/lowest low for x bars. Not very good - but maybe I can tweek it a bit.

Let's see if we can get this really cooking and reduct the drawdowns.

B

[dloomis]

Re: Reply to post 'Signal clarity'

> Doji3333 has just replied.
>
> Here are the contents of the post:
> [quote]
> Hey Guys...

Wow, thanks for the input, no problem here with any help we get.

I think a little tweeking here and there by Doji and some by FG and maybe we
will have two good systems going.

Let's be sure we keep track of which system is which so any lurkers won't
get any more confused than I am getting.

Do we need a list of the current directions we are going in?

Can we use the subject to describe the system we are refering to in the
post?

----- Original Message -----
From: <[email protected]>
To: <[email protected]>
Sent: Tuesday, December 16, 2003 4:24 PM
Subject: Reply to post 'Signal clarity'


> Hello dloomis,
>
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>

[Doji3333]

Organization...

David,

I'll post my files into a directory on your fileshare group. That way, everything is separate from yours.

Brad

[ACM]

FWIW so far this is the best result I have come up with using the raw system. Settings can be seen in the indicator pane
Alex

[dloomis]

Re: Reply to post 'Signal clarity'

Alexis

Can you describe how you came up with those settings? Is that 190 pts or
$190.

[FibbGann]

I am currently righting down a few ways to to accomplish or objective. When I am done I will post them. And you all can say what you think.

Fibbgann

[ACM]

David
I believe it is points
Alex

[ACM]

With a slight change in the strategy (still in its raw form) and without too much testing here are some results together with the equity curve. The parameters for the attached efs need to be set to 28/14/16-3 respectively for the CCI, RSI and Stoch. Chart settings are 3 minute, 9:30-16:15 and 120 days.
The Back Tester Report is available here in a zip file
Alex

[dloomis]

Re: Reply to post 'Signal clarity'

Nice work! Nice equity line!!

Any reason why the equity curves aren't present in the zip file?


----- Original Message -----
From: <[email protected]>
To: <[email protected]>
Sent: Wednesday, December 17, 2003 5:22 PM
Subject: Reply to post 'Signal clarity'


> Hello dloomis,
>
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>

[ACM]

David
I don't know why but it does not save the graphs in html.
Alex