|
PureBytes Links
Trading Reference Links
|
<x-html><!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD><TITLE></TITLE>
<META content="text/html; charset=iso-8859-1" http-equiv=Content-Type>
<META content="MSHTML 5.00.2014.210" name=GENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=#ffffff>
<DIV><FONT color=#000080 face=Arial><SPAN
class=250160302-16061999>Walter,</SPAN></FONT></DIV>
<DIV><FONT color=#000080 face=Arial><SPAN class=250160302-16061999> The
only thing I'd be beware of is which ATR type and period she
used.</SPAN></FONT></DIV>
<DIV><FONT color=#000080 face=Arial><SPAN
class=250160302-16061999></SPAN></FONT> </DIV>
<DIV><FONT color=#000080 face=Arial><SPAN class=250160302-16061999>The built in
Wilder ATR uses 'his EMA', so the period is long [ 14 gives a true EMA of
27].</SPAN></FONT></DIV>
<DIV><FONT color=#000080 face=Arial><SPAN
class=250160302-16061999></SPAN></FONT> </DIV>
<DIV><FONT color=#000080 face=Arial><SPAN class=250160302-16061999>Chandre used
a SMA of the TR for volatility-based stops, and I'd bet that is what
Brown uses.</SPAN></FONT></DIV>
<DIV><FONT color=#000080 face=Arial><SPAN
class=250160302-16061999></SPAN></FONT> </DIV>
<DIV><FONT color=#000080 face=Arial><SPAN class=250160302-16061999>A custom
formula for the ATR (SMA or EMA) is</SPAN></FONT></DIV>
<DIV><FONT color=#000080 face=Arial><SPAN
class=250160302-16061999>Len:=Input("Periods",1,400,89);<BR>(Mov((H - L +
Abs(H - Ref(C,-1)) + Abs(L - Ref(C,-1)) ),len,E))/2 <BR></DIV></SPAN></FONT>
<DIV><FONT color=#000080 face=Arial><SPAN
class=250160302-16061999></SPAN></FONT> </DIV>
<DIV><FONT color=#000080 face=Arial><SPAN
class=250160302-16061999>Bob</SPAN></FONT></DIV>
<DIV><FONT color=#000000 face=Arial size=2><A
href="mailto:bjagow@xxxxxxx";>bjagow@xxxxxxx</A></FONT></DIV>
<DIV> </DIV>
<BLOCKQUOTE style="MARGIN-RIGHT: 0px">
<DIV class=OutlookMessageHeader><FONT face="Times New Roman"
size=2>-----Original Message-----<BR><B>From:</B>
owner-metastock@xxxxxxxxxxxxx [mailto:owner-metastock@xxxxxxxxxxxxx]<B>On
Behalf Of</B> Walter Lake<BR><B>Sent:</B> Tuesday, June 15, 1999 6:34
PM<BR><B>To:</B> metastock@xxxxxxxxxxxxx<BR><B>Subject:</B> Re: Volatility
bands on Osc's<BR><BR></DIV></FONT>
<DIV><FONT size=2>Hi Bob</FONT></DIV>
<DIV> </DIV>
<DIV><FONT size=2>You're absolutely right.</FONT></DIV>
<DIV> </DIV>
<DIV><FONT size=2>Any suggestions on programming Brown's volatility bands into
either Excel or Metastock?</FONT></DIV>
<DIV> </DIV>
<DIV><FONT size=2>==============</FONT></DIV>
<DIV> </DIV>
<DIV><FONT size=2>{Is there a range of coefficients given in her book?
The<BR>CoefficientDown(2.1) and CoefficientUp(2.3) get inserted into the
formulas<BR>below, just like in
Metastock.}<BR><BR>Plot1[[Average[[RSI[Close,14]],6]]+[Coefup*[Average[TrueRangeCustom[[RSI[Clo<BR>se<BR>,14]],[RSI[Close,14]],[RSI[Close14]]],15]]],"Plot1"];<BR><BR>{The
code separates True Range into 2 parts: first the True Range then
the<BR>Average. Whereas in Metastock you only use ATR. What is the input for
the<BR>average and the ATR? 6 and
15?}<BR><BR>Plot2[[Average[[RSI[Close,14]],6]]-[Coefdwn*[Average[TrueRangeCustom[[RSI[Cl<BR>os<BR>e,14]],[RSI[Close,14]],[RSI[Close,14]]],15]]],"Plot2"];<BR><BR>{As
Adam suggested, the ATR will probably be calculated on the price and<BR>then
applied to the RSI
indicator}<BR><BR>Plot3[[RSI[Close,14]],"Plot3"];<BR><BR>=======</FONT></DIV>
<DIV> </DIV>
<DIV><FONT size=2>Best regards</FONT></DIV>
<DIV> </DIV>
<DIV><FONT size=2>Walter</FONT></DIV>
<BLOCKQUOTE
style="BORDER-LEFT: #000000 2px solid; MARGIN-LEFT: 5px; MARGIN-RIGHT: 0px; PADDING-LEFT: 5px; PADDING-RIGHT: 0px">
<DIV style="FONT: 10pt arial">----- Original Message ----- </DIV>
<DIV
style="BACKGROUND: #e4e4e4; FONT: 10pt arial; font-color: black"><B>From:</B>
Bob Jagow </DIV>
<DIV style="FONT: 10pt arial"><B>To:</B> <A
href="mailto:metastock@xxxxxxxxxxxxx";
title=metastock@xxxxxxxxxxxxx>metastock@xxxxxxxxxxxxx</A> </DIV>
<DIV style="FONT: 10pt arial"><B>Sent:</B> Tuesday, June 15, 1999 3:45
PM</DIV>
<DIV style="FONT: 10pt arial"><B>Subject:</B> RE: Volatility bands on
Osc's</DIV>
<DIV><BR></DIV>
<DIV><FONT color=#000080 face=Arial>Walter,</FONT></DIV>
<DIV><FONT color=#000080 face=Arial> My Kaufman shows</FONT></DIV>
<DIV><FONT color=#000080 face=Arial> 2) H today - C
yesterday <FONT color=#ff0000>not</FONT></FONT></DIV>
<DIV><FONT color=#ff0000 face=Arial> 2) H today - L yesterday
<BR></FONT></DIV>
<DIV><FONT color=#000080 face=Arial>With that change, all three of those
defs are identical, leaving OT alone in left field.</FONT></DIV>
<DIV><FONT color=#000080 face=Arial></FONT> </DIV>
<DIV><FONT color=#000080 face=Arial>Bob<BR><BR></FONT><FONT
face=Arial><BR><BR></DIV></FONT>
<P><FONT size=2><FONT face=Arial size=3><A
href="mailto:bjagow@xxxxxxx";>bjagow@xxxxxxx</A> <<A
href="mailto:bjagow@xxxxxxx";
target=_blank>mailto:bjagow@xxxxxxx</A>><BR><BR><BR>-----Original
Message-----<BR>From: <A
href="mailto:owner-metastock@xxxxxxxxxxxxx";>owner-metastock@xxxxxxxxxxxxx</A><BR>[<A
href="mailto:owner-metastock@xxxxxxxxxxxxx";
target=_blank>mailto:owner-metastock@xxxxxxxxxxxxx</A>]On Behalf Of Walter
Lake<BR>Sent: Tuesday, June 15, 1999 11:54 AM<BR>To:
metastock@xxxxxxxxxxxxx<BR>Subject: Re: Volatility bands on
Osc's<BR><BR><BR>Hi Bob<BR><BR>Yep ... Page 82 of the OT systems book. This
was part of the system that OT<BR>sold as an insert into
MetaStock.<BR><BR>Anyway ... here are three more definitions of True
Range<BR><BR>Kaufman refers to Wilder's DM or Directional Movement as True
Range, i.e.<BR>the larger of<BR>1) H today - L today<BR>2) H today - L
yesterday <BR>3) C yesterday - L today<BR><BR><BR>>From the Metastock
Custom formula collection: The Volatility Index (VI) is<BR>described by
Wilder as:<BR>VI Today = (13 * VI Prev + TR1) / 14 *where TR1 is today's
true range.<BR>He defines the true range as the greatest of the
following:<BR>The distance from today's high to today's low<BR>The distance
from yesterday's close to today's high, or<BR>The distance from yesterday's
close to today's low.<BR><BR><BR>Schwager's True Range = True High less True
Low<BR>True High = max of H today and the previous days C (whichever is
greater)<BR>True Low = min of L today and previous days C<BR>(whichever is
less)<BR><BR>Best regards<BR><BR>Walter<BR><BR>-</FONT>---- Original Message
-----<BR>From: Bob Jagow <bjagow@xxxxxxx><BR>To:
<metastock@xxxxxxxxxxxxx><BR>Sent: Tuesday, June 15, 1999 12:03
PM<BR>Subject: RE: Volatility bands on Osc's<BR><BR><BR>> TR isn't based
on the open, Walter.<BR>> Did OT publish that? No wonder it
makes such amazing calls :o)<BR>><BR>> bjagow@xxxxxxx <<A
href="mailto:bjagow@xxxxxxx";
target=_blank>mailto:bjagow@xxxxxxx</A>><BR>><BR>><BR>>
-----Original Message-----<BR>> From:
owner-metastock@xxxxxxxxxxxxx<BR>> [<A
href="mailto:owner-metastock@xxxxxxxxxxxxx";
target=_blank>mailto:owner-metastock@xxxxxxxxxxxxx</A>]On Behalf Of Walter
Lake<BR>> Sent: Tuesday, June 15, 1999 7:33 AM<BR>> To:
metastock@xxxxxxxxxxxxx<BR>> Subject: Re: Volatility bands on
Osc's<BR>><BR>><BR>> Adam, Lenny, Laurent, et all<BR>><BR>>
Here's the code for the Omni Trader Volatility Breakout System. It
has<BR>True<BR>> Range separated from the Average.<BR>><BR>> "...
the following calculations are made to arrive at the
Volatility<BR>Bands:<BR>><BR>> If<BR>> P = Periods used<BR>>
Range = Breakout Percentage (that is, 150 = 150%)<BR>> Then<BR>> R1 =
ABS(O today -C today)<BR>> R2 = ABS(O today - C yesterday)<BR>> TR =
Greater of R1, R2 {True Range calculation)<BR>> AVGTR = Average of TR of
P periods<BR>><BR><BR></FONT></P></BLOCKQUOTE></BLOCKQUOTE></BODY></HTML></x-html>From ???@??? Wed Jun 16 07:13:43 1999
Received: from listserv.equis.com (204.246.137.2)
by mail02.rapidsite.net (RS ver 1.0.4) with SMTP id 027384;
Wed, 16 Jun 1999 06:02:28 -0400 (EDT)
Received: (from majordom@xxxxxxxxx)
by listserv.equis.com (8.8.7/8.8.7) id DAA04652
for metastock-outgoing; Wed, 16 Jun 1999 03:45:19 -0600
X-Authentication-Warning: listserv.equis.com: majordom set sender to owner-metastock@xxxxxxxxxxxxx using -f
Received: from freeze.metastock.com (freeze.metastock.com [204.246.137.5])
by listserv.equis.com (8.8.7/8.8.7) with ESMTP id DAA04649
for <metastock@xxxxxxxxxxxxxxxxxx>; Wed, 16 Jun 1999 03:45:16 -0600
Received: from mail.apsydev.com (root@xxxxxxxxxxxxxxxxxxxx [62.160.31.1])
by freeze.metastock.com (8.8.5/8.8.5) with ESMTP id DAA03770
for <metastock@xxxxxxxxxxxxx>; Wed, 16 Jun 1999 03:58:44 -0600 (MDT)
Received: from lgt (lgt.apsydev.com [62.160.31.54])
by mail.apsydev.com (8.8.7/8.8.7) with SMTP id LAA06710
for <metastock@xxxxxxxxxxxxx>; Wed, 16 Jun 1999 11:49:17 +0200
From: "Laurent GITTLER" <lgittler@xxxxxxxxxxx>
To: <metastock@xxxxxxxxxxxxx>
Subject: RE: Off Topic: Choice of Processor
Date: Wed, 16 Jun 1999 11:47:04 +0200
Message-ID: <001701beb7dd$306eb650$361fa03e@xxxxxxxxxxx>
MIME-Version: 1.0
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3 (Normal)
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook 8.5, Build 4.71.2173.0
X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2014.211
Importance: Normal
In-Reply-To: <01BEB725.C80694A0.ay286@xxxxxxxx>
Sender: owner-metastock@xxxxxxxxxxxxx
Precedence: bulk
Reply-To: metastock@xxxxxxxxxxxxx
X-Loop-Detect: 1
X-UIDL: ec4e769320c6ee2ccc1c06cba405ace5.07
The dis imaging program I use is powerquest DiskImage 2.0 it is not
expensive (approx $70), and works fine. You don't need the professional
version of DiskImage.
I am not running it live on NT (it does not work), but on DOS so my cache is
not active when I make a disk image. To make a disk image, I boot on DOS and
I run the program. It is anyway easier to use on Win98, since it switches
automatically from Win98 to DOS without rebooting. I choose to make a disk
image on the second IDE drive. It is fast and reliable. It is even possible
to automate the backup process through scripts. Personally I zip all my
stock data files on the second disk every day (actually I have an 640 Mb
optical drive on which I store zipped files as well) and I backup my
computer with disk image once a every week.
I know they are other from Symantec and formerly Quaterdeck (now Symantec),
and others.
To make a copy, you can choose to make a disk copy (i.e. the first disk is
duplicated on the second one and if the first disk fails, all you have to do
is only to replace the main disk with the second one and your system is up
again within minutes) or you can choose to store the copy on a file on the
second disk, and therefore, mae images of the first disk and store them on
the second disk (if the second disk is large enough, you can store several
disk images).
When considering a buy, I would say either 1) buy a system with a really
fast main disk and either a larger but slower secondary disk to store
several disk images on it (my personal preference), or 2) buy equally large
hard disks and put the second one in a mobile rack and mount it for disk
images, then put it off the system.
In the first case, don't forget to have a dos bootable floppy disk and disk
image to restore the disk when it crashes.
>From what I experimented, what my advice to choose disk image rather then
disk copy if you keep both hard drives online, because you will never get
confused with the Windows explorer between partitions one having the real
data, and the second also visible from the windows explorer having a copy of
the first one. You can easily get confused make mistakes.
What I would would even recommend is to keep a partition empty on the second
disk which has the same size as the partition on the main disk (and keep it
empty), then additional extended partitions to store images in. If you have
a disk crash (about 75% of failures I saw on computers where I work are
coming from disks which crashed and was totally lost when the weather was
getting hot), then from the mother board or graphic card. It is scarce that
CPU or memory fails.
I also saw PC crashes with problem coming from the mains, when voltage was
peaking too high, so it is important to use a UPS to prevent that.
According to me, this is the fastest, reliable and less expensive way to
make backups. A second hard disk is about the same price as a tape backup
(or even cheapest) and far more faster and reliable.
A third possibility which works on Windows NT when you have 2 exact same
disks is to make a soft disk mirror with no addtional software, just a
driver service. You have live copy in real time (data is stored on both two
disks rather than on one disk, but I didn't choose it because I want to
protect myself from both disk hardare failure and operating system failure
or PC crash leading to an unrecoverable error in system FAT or NTFS). So I
think it is safer to make hard disk backups, with a small loss of data
penalty between two backups if the system crashes, rather than trusting the
operating system. That's a choice.
I mention it, because if you want to invest a little more on your computer,
you can choose additional security using 3 disks (1 main disk + mirror), and
one auxiliary disk dedicated to backups.
The url for the disk image sofware is www.powerquest.com. They have a trial
version.
V2 works fine it is $69.95.
-----Original Message-----
From: owner-metastock@xxxxxxxxxxxxx
[mailto:owner-metastock@xxxxxxxxxxxxx]On Behalf Of John Sellers
Sent: Tuesday, June 15, 1999 8:40 PM
To: 'metastock@xxxxxxxxxxxxx'
Cc: 'Rich Bulow'
Subject: RE: Off Topic: Choice of Processor
May I ask what name of the software you use for disk imaging? The fact that
you have disk cache may be the reason you can back up in such a small
amount of time as 5 minutes. I do not have disk cache and am using the win
98 back up program from hard drive to another at about 40MB per minute thus
I need at least 25 minutes for backing up 1 GB of data and program
settings. However I find this back up software coming with Win 98 namely
written by Seagate has shortcomings and is not reliable in my opinion.
If you may please give us your thoughts and recommendations about improving
performance using Win 98. I assume that the cache program is not available.
It is possible with sufficient RAM to make a RAM disk and use it for
locating the swapping disk. This thought leads me to expect a much faster
data transfer if one implements this approach.
-----Original Message-----
From: Laurent GITTLER [SMTP:lgittler@xxxxxxxxxxx]
Sent: Tuesday, June 15, 1999 5:33 AM
To: metastock@xxxxxxxxxxxxx
Subject: RE: Off Topic: Choice of Processor
The NT cache or Win 95/98 is a write-through cache, that means each time
there is a modification to be written on a file, the write must be done on
the Hard Disk before the program can continue to do other tasks.
This has advantages : Data is theorically never lost on disk, I say
theorically because practically, it might be lost when there is a power
failure and the file was not closed.
But it has also a huge impact on overall performance. SuperCache NT from
eecsys (I have never seen another program doing that on NT/ Win 95/98, but
they might be some) can be configured as a write-back cache : That means
that data write is commited before it is actually written to the disk, it
will be written on background by a lazy process, approximately 1 to 5
seconds later. This leads to a tremendous performance enhancement of the
overall system.
There is one big disadvantage as well if you are not protected against
power
failure, a lot of data might be lost (up to 10 seconds). Therefore, make
sure you use an UPS before activating a write-back disk cache, and
regularly do backups.
Anyway, for me, there is definetly no confidence in a PC without power
failure. data might be lost anyway, so there is not really a big difference
between write back and write through cache, if the system crashes due to a
power failure, there is a high risk you loose data anyway.
There might be also some specific hardware to do this, this is very common
on large computers like SUN enterprise 5000 and other unix serve boxes on
which the hard disk controller has a large disk cache backed up by a
battery
able to keep data up to 1 year when power is lost, and data will be written
on disk as soon as power is back.
This technology is also available on PC as well, on RAID controllers like
MYLEX DAC 960 or DPT or other Hard disk controllers, it is more expensive
and requires ultra wide SCSI disks, the complete hard disk system price
will
be around $3000-$4000, but can give access to other possible performance
enhancement like raid 0 (data stripping i.e. if you are using 4 disks, then
all 4 disks will be seen as one single large volume and data written will
be
splitted on disks by blocks of 8 K, it is used mainly for SQL databases and
programs on which performance is critical. I do not believe it would worth
the price unless you have a very large portofolio on which one day off
would
result in a potential large loss. But in that case, I believe the smartest
thing would be to have a spare PC used as backup of the regular PC.
For backups : personally as I upgraded my PC, I changed my disk for a
larger
one and I used my old disk as a backup disk on which I store disk images.
It
is really faster than a tape backup (approx 5 minutes to do the backup of
the full disk).
-----Original Message-----
From: owner-metastock@xxxxxxxxxxxxx
[mailto:owner-metastock@xxxxxxxxxxxxx]On Behalf Of John Sellers
Sent: Monday, June 14, 1999 6:49 PM
To: 'metastock@xxxxxxxxxxxxx'
Subject: RE: Off Topic: Choice of Processor
This approach sounds interesting but would you please explain more about
the cache. I am aware that my computer has a swapping file which is managed
in size by the Windows system program or may be set to some amount of fixed
memory for each of the minimum and maximum values. I am aware that setting
both max and min values to one value a sufficiently high value for ones
programs; this approach I found from my experience achieved an increase in
operationally speed. I believe it may have been accomplished because of
faster data transfers.
Is this cache memory something different? I believe all of our group should
comprehend computer functionality because most of us at some time in the
future shall be considering purchase of an improved computer .
-----Original Message-----
From: Laurent GITTLER [SMTP:lgittler@xxxxxxxxxxx]
Sent: Monday, June 14, 1999 8:16 AM
To: metastock@xxxxxxxxxxxxx
Subject: RE: Off Topic: Choice of Processor
Here is my personal thoughts, it won't be what people will generally say,
but it is quite easy to find out if you are using several OS and compare
them, ie. Win 95&NT vs. Linux).
I am not using MS PRO, and what I do are genrally explorations on 9000
stocks to detect some patterns, divergence and so on on all stocks. What I
want is those explorations to be run as fast as possible, and I want smotth
screen output as well.
What I found is is somehow amazing : The CPU power on WIN 95/NT is not the
key element of the overall PC performance. Memory and hard disk have the
most impact on response time on explorations.
I have a K6-233 I bought a year and a half ago (so said "old CPU"), NT4
workstation. What I monitored is bottleneck is made on disk activity while
exploring. I switched from an average exploration time of 17 minutes to 6
minutes just by using a good hard disk write behind cache program. I made
test on a PII 350 MHz, NT4, 256 Mb RAM without hard disk write back cache,
and again, I got explorations of 16 minutes.
More than 32 Mb of my 96 Mb Memory are dedicated to cache. I consider
upgrading my system to a Celeron 350 MHz or K6-2 350 or whatever, but
emphasize the PC upgrade on memory (256 K RAM or 512 Mb RAM) and real fast
hard disks.
I wish I had 1 Gb RAM and then I would do MS explorations in RAM disk with
background writes to the Hard disk, I am sure it will really give blasting
performance.
So my advise would be : make sure you have the best hard disk, or even
consider RAID 0 disks (stripping) with write back cache. But all I say is
for explorations (i.e regularly scan the databse for patterns and so on).
My
problem might be totally different from yours.
Laurent GITTLER
-----Original Message-----
From: owner-metastock@xxxxxxxxxxxxx
[mailto:owner-metastock@xxxxxxxxxxxxx]On Behalf Of Essan Soobratty
Sent: Tuesday, June 15, 1999 3:36 PM
To: metastock@xxxxxxxxxxxxx
Subject: Off Topic: Choice of Processor
Hi,
I am about to upgrade my current PC which is an NEC PII 233mhz. I am
not an expert when t comes to hardware and therefore seek views as to an
appropriate upgrade.
Its main use would be for running metastock pro, excel, word, netscape
etc and it would be part of a LAN network. I don't think I would be
using it for any heavy duty graphics.
Browsing the shelves of Comp-USA it looks like my choices are between an
Intel Celeron, PII or PIII. Can anyone give some suggestions and in
particular describe the differences between the 3 processors? Whatever
the choice, I would have a minimum of 64meg or RAM and would prefer to
use win95.
Also as a bechmark, which Celeron chip would be the equivalent to my
current PII 233mhz?
Email direct if you prefer. Thanks in advance.
Essan Soobratty.
|