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

[RT] S&P 500 Neural Net Predictions - Questions addressed



PureBytes Links

Trading Reference Links

<x-html><!DOCTYPE HTML PUBLIC "-//W3C//DTD W3 HTML//EN">
<HTML>
<HEAD>

<META content=text/html;charset=iso-8859-1 http-equiv=Content-Type>
<META content='"MSHTML 4.72.3616.1301"' name=GENERATOR>
</HEAD>
<BODY bgColor=#ffffff>
<DIV><FONT color=#000000 size=2>Based on several of the comments and questions I 
received based on my post yesterday, I'm going to attempt to address as many 
questions as I can in this post.&nbsp; I use a product called Profit by BioComp 
Systems to develop my models.&nbsp; Visit <A 
href="http://www.biocompsystems.com";>http://www.biocompsystems.com</A> if you 
want to learn more about their products.&nbsp; I'm in no way affiliated with 
them except that I have been a customer of theirs for about one year and now use 
their neural products in my work as well (non financial aplications).&nbsp; 
</FONT><FONT color=#000000 size=2>My models are in the market 100% of the time 
and trade on the following day's open based on neural signals generated the 
night before.&nbsp; The neural signals are updated each night with the current 
day's data.&nbsp; The models do not try to pick each up and down day, rather 
they try to capture short-term market trends.&nbsp; This is more a function of 
my modeled output than anything.&nbsp; The models assume $100 per trade for 
commissions and slippage.&nbsp; The <FONT color=#008000 face=Arial size=2><FONT 
color=#008000 face=Arial size=2><STRONG>green</STRONG></FONT> </FONT>dots on the 
price chart indicate a day when the model shifted from a short to a long 
position.&nbsp; Conversely, a <FONT color=#ff0000 face=Arial size=2><STRONG>red 
</STRONG></FONT>dot indicates a day where the model shifted from a long to a 
short position at that day's opening price.&nbsp; The <FONT color=#000000 
face=Arial size=2>yellow </FONT>line running through the price data is nothing 
more than a simple 5-day moving average (SMA) of the closing prices.&nbsp; I 
added that as a visual tool for myself and it has nothing to do with the models 
themselves (other than the fact that some of the 5 system models use the daily 
change in the 5-day SMA as a model input).&nbsp; The output of the neural model 
is shown in the middle section of the chart.&nbsp; It appears as an osillator 
moving between -1.0 and +1.0.&nbsp; A new long (short) position is indicated 
when the signal moves from below (above) the zero (0) line to above (below) it 
(this is when you see a green (red) dot at the next day's open). The system 
model, which I will post shortly, has a signal that oscillates between -5 and +5 
since there are 5 total models and the system model uses a voting process to 
generate its signal.&nbsp; The system model's signal doesn't look like the 
posted model's sine wave pattern, but the initiation of a long or short position 
is generated the same way (crossing the zero line).&nbsp; The system model may 
stay at zero for several days, but only when it <U><FONT color=#000000 
face=Arial size=2>crosses</FONT></U> the zero line is a new position 
taken.&nbsp; I do not plan on posting the models' updated signals each day other 
than for my own trading.&nbsp; My hope is that this will spark an interest in 
others to look into neural nets as a possible trading aid.&nbsp; My apologies 
for referring to my &quot;web page.&quot;&nbsp; Since I'm not into generating 
web pages as we all know and love them, I will simply be posting nonlinked gifs. 
Hope this at least addresses most of the responses I received. - 
Brian</FONT></DIV></BODY></HTML>
</x-html>From ???@??? Fri Jan 21 18:11:12 2000
Return-Path: <listmanager@xxxxxxxxxxxxxxx>
Received: from mail.thetrellis.net ([208.179.56.11])
	by purebytes.com (8.8.7/8.8.7) with SMTP id TAA18445
	for <neal@xxxxxxxxxxxxx>; Fri, 21 Jan 2000 19:02:41 -0800
Received: from REALTRADERS.COM
	([208.179.56.198])
	by mail.thetrellis.net; Fri, 21 Jan 2000 17:56:50 -0800
Received: from crusoe.crusoe.net by realtraders.com
	with SMTP (MDaemon.v2.8.5.0.R)
	for <realtraders@xxxxxxxxxxxxxxx>; Fri, 21 Jan 2000 17:51:23 +0000
Received: from pgms ([206.136.71.165]) by crusoe.crusoe.net
          (Post.Office MTA v3.5.3 release 223 ID# 0-52487U2500L250S0V35)
          with SMTP id net; Fri, 21 Jan 2000 20:53:48 -0500
Message-ID: <014f01bf647b$8496e5e0$a34788ce@xxxx>
Reply-To: "swp" <swp@xxxxxxxxxx>
From: "swp" <swp@xxxxxxxxxx>
To: "<realtraders@xxxxxxxxxxxxxxx>"
	<realtraders@xxxxxxxxxxxxxxx>
References: <NDBBJDGFKDOOCOMFKPBHMEBICDAA.mknapp@xxxxxxxxxxxxxx>
Subject: [RT] Re: GEN: YIELD CURVE
Date: Fri, 21 Jan 2000 20:53:44 -0500
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.00.2314.1300
X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2314.1300
X-MDaemon-Deliver-To: realtraders@xxxxxxxxxxxxxxx
X-Return-Path: swp@xxxxxxxxxx
Sender: listmanager@xxxxxxxxxxxxxxx
X-MDMailing-List: realtraders@xxxxxxxxxxxxxxx
X-MDSend-Notifications-To: listmanager@xxxxxxxxxxxxxxx
Status:   

Hang on a second. 3-year bills are not above the 30Y! It is not even
close. The 10Y now trades over bonds, and the 5Y is nearby. But
remember, there is also a lot less supply of the 30Y these days.

---
Steven W. Poser, President
Poser Global Market Strategies Inc.

url: http://www.poserglobal.com
email: swp@xxxxxxxxxxxxxxx

Tel: 201-995-0845
Fax: 201-995-0846
----- Original Message -----
From: <mknapp@xxxxxxxxxxxxxx>
To: <realtraders@xxxxxxxxxxxxxxx>
Sent: Friday, January 21, 2000 7:12 PM
Subject: [RT] Re: GEN: YIELD CURVE


> RT-er's,
>
> I am not ruling out a recession, but their may be other reasons as to
why
> the yield curve has inverted.  A decent contributor to this particular
> occurrence seems to be supply.  The treasuries buy back program is
> inadequate and the plan seems to have let fixed income players know
that the
> treasury has no credible way of dealing with the supply issue.  The
Fed's
> clarification of their bias statement did nothing to quell fears of
them
> falling "behind the curve" either.  The increase in the 10y t-note
> volatility could be a confirmation of this point.  The flattening
could be
> entirely explained away by these issues.  I personally don't buy into
> inflation, but the stock market and oil is getting harder and harder
to
> ignore.  Of course, only time will tell what is really going on.
>
> Trade Well
>
>  > RTs:
> | >
> | > I have a note from several years ago that states "when IRX > ITX
there
> is
> | > always a recession, and there has never been a false signal. I
have no
> | idea
> | > of the truth of this statement, but it was just discussed as
credible
> on
> | > CNBC. I agnostic on this subject, but curious how index and option
> traders
> | > might consider this.
> | >
> | > Regards,
> | >
> | > Michael
> | >
>
>
>