|
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 4.72.2106.6"' name=GENERATOR>
</HEAD>
<BODY bgColor=#ffffff>
<DIV><FONT color=#000000 size=2>Good Morning RealTraders:</FONT></DIV>
<DIV><FONT color=#000000 size=2></FONT> </DIV>
<DIV><FONT color=#000000 size=2>Well it sure has been an interesting week with
the Galaxy 4 satellite going out of orbit. This has caused a real sore spot for
a lot of us and in a lot of different ways. I am going to dispense with the
reasons and whining and go right into the future.</FONT></DIV>
<DIV><FONT color=#000000 size=2></FONT><FONT
size=2>
<STRONG>DISASTER RECOVERY PLANNING</STRONG></FONT></DIV>
<DIV><FONT size=2>Satellites have been extremely reliable in the past, but not
without flaws. Technology is not perfect and subject to human error. Knowing our
limits should become one of our greatest strengths. We must not place all
the blame on the data vendors but some on ourselves perhaps. <FONT size=2>This
is where our minds should be focusing right now. If we are responsible traders,
we should have a written disaster recovery plan for such circumstances first and
foremost. Secondly, we should also take the time to formally write our data
vendors and request they do the same. Some of them may say, we have one already,
but it is more than obvious, it is inadequate based on the amount of downtime.
Disaster Recovery Plans are like due diligence, an ongoing process that requires
continual review and updating. </FONT></FONT></DIV>
<DIV><FONT color=#000000 size=2></FONT> </DIV>
<DIV><FONT color=#000000 size=2><STRONG>My plea to all
RealTraders</STRONG>:</FONT></DIV>
<DIV><FONT color=#000000 size=2></FONT> </DIV>
<DIV><FONT color=#000000 size=2>Later this week I will post the name and
addresses of the President's of each data vendor that I can find. I would like
to ask all of you to write them and request that they review their disaster
recovery plans, or establish one if they don't have one. The idea is to work
with the data vendors in a constructive manner so as to avoid such occurrences
in the future. If I were them, I would view this situation as a great marketing
opportunity by touting a comprehensive disaster recovery plan. The objective is
zero downtime. Although this may be deemed unrealistic, it should be a
goal. </FONT></DIV>
<DIV><FONT color=#000000 size=2></FONT> </DIV>
<DIV><FONT color=#000000 size=2>Respectfully,</FONT></DIV>
<DIV><FONT color=#000000 size=2></FONT> </DIV>
<DIV><FONT color=#000000 size=2>Richard J. Chehovin</FONT></DIV></BODY></HTML>
</x-html>From ???@??? Mon May 25 11:56:50 1998
X-POP3-Rcpt: neal@xxxxxxxxx
Return-Path: <owner-realtraders@xxxxxxxxxxxxxx>
Received: from accessone.com (list.listserver.com [198.68.191.15])
by purebytes.com (8.8.4/8.8.4) with ESMTP
id KAA09683 for <neal@xxxxxxxxxxxxx>; Mon, 25 May 1998 10:52:26 -0700
Received: from localhost (localhost [127.0.0.1])
by accessone.com (8.8.5/8.8.5/PIH) with SMTP id LAA24665;
Mon, 25 May 1998 11:39:48 -0700 (PDT)
Received: from intrepid.intrepid.com (root@xxxxxxxxxxxxxxxxxxxxx [192.195.190.1])
by accessone.com (8.8.5/8.8.5/PIH) with ESMTP id LAA24447
for <realtraders@xxxxxxxxxxxxxx>; Mon, 25 May 1998 11:36:26 -0700 (PDT)
Received: from presto.intrepid.Com (presto [192.195.190.8])
by intrepid.intrepid.com (8.8.7/8.8.7) with ESMTP id KAA16244
for <realtraders@xxxxxxxxxxxxxx>; Mon, 25 May 1998 10:39:02 -0700
Received: by presto.intrepid.Com (8.7.5/3.2.083191-Intrepid Technology)
id SAA06652; Mon, 25 May 1998 18:35:36 GMT
Message-Id: <9805251135.ZM6648@xxxxxxxxxxxxxxxxxxx>
Date: Mon, 25 May 1998 11:35:35 -0700
Reply-To: gary@xxxxxxxxxxxx
Sender: owner-realtraders@xxxxxxxxxxxxxx
From: gary@xxxxxxxxxxxx (Gary Funck)
To: RealTraders Discussion Group <realtraders@xxxxxxxxxxxxxx>
Subject: MKT - OEX v. VIX
In-Reply-To: THE DOCTOR <droex@xxxxxxxxxxxx>
"Re: MKT - VIX LEVELS" (May 24, 1:26pm)
References: <3566C5A8.2F1B4AD3@xxxxxxxxxxxx>
<9805240931.ZM4495@xxxxxxxxxxxxxxxxxxx>
Mime-Version: 1.0
Content-Type: multipart/mixed;
boundary="PART-BOUNDARY=.19805251135.ZM6648.Intrepid.Com"
X-Mailer: Z-Mail (3.2.3 08feb96 MediaMail)
X-Listprocessor-Version: 8.1 -- ListProcessor(tm) by CREN
Attached is a (rather busy) chart showing OEX in the top plot, VIX
in the second plot, the correlation between 5 day changes in VIX
and subsequent 5 day close-to-close (log) changes in OEX (except
VIX is differenced by subtracting the older value from the value
5 bars later), and in the bottom chart (in yellow), a look at the
"tracking erorr" of VIX relative to actual volatility.
In the top chart of OEX, certain bars are painted blue and red
to note divergences from what I will call the normal scenario:
- when VIX falls, OEX typically rises
- when VIX rises, OEX typically falls
The blue bars mark points at which the 5-day OEX close-to-close change
is positive, and the 5 day change in VIX is also positive. The
red bars mark points at which the 5-day close-to-close change in
OEX is negative, and so is the 5-day change in VIX. As you can
see thsse divergences aren't that common, occurring maybe 1/3 of
the time, and the "complacent bull" (red) bars are less
frequent than the "worried bull" (blue) bars. You'll note that
recently we had 4 bars in a row where the 5-day OEX momentum was
negative, and the 5 day momentum in VIX was also negative.
The red bars are indicative of the times when the bulls seem to be
looking the other way, and even though the market is declining,
are not pushing VIX up (presumably by buying protective puts).
In 5 out of the 7 cases shown, the market was generally up 3 days
later, following a red bar. One noteable exception was October-97.
In general, these red bars often appeared closer to the the bottom
of a move than a top, and if you'd bought the close of a red bar,
using that bar's low as a stop, and held for 5 days this was
usually a winning trade.
The blue bars (woorried bulls) occur when both OEX is climbing
and so is VIX. They aren't as easy to classify, except they seem
to crop up towards the end of a strong up move, as it consolidates,
or on the up-leg in a V-bottom, where the high risk buyers begin
to step in.
The second, purple plot shows VIX on a semi-log chart.
The red plot shows the correlation between 5-day changes in VIX
and the percent (actually log) change in OEX the following 5-days,
using a 10-sample interval. The sense of VIX is inverted (the
older value is subtracted from the newer value), because typically
a down move in VIX precedes and up move in OEX and vice versa.
Note that the correlation values are strong (the blue line is at
0.500), spending most of its time above 0.5, showing this relationship
is rather consistent. The times where this line dips below 0.5,
are when there's a divergence. Late March-98 is one the stronger
divergences ... note that the grouping of blue bars in that time
frame is in sync. with the correlation diving below 0 (as expected).
The bottom chart shows the difference betwwen VIX 5 days ago, and
the actaul historic/statistical volatility over the past 10 days.
This differnce is measured as a percentage of VIX, not as a straight
difference.
I'm using the following to calculate statistical volatility:
VLTY = 100 * AvgTrueRange(10)/Close * SQRT(260)
In general, VIX seems to trade at about 10% above that level.
The blue line is drawn at the 0% level. When the yellow line
is above the blue line, VIX is "expensive" relative to its normal
levels, and is overcompensating in its extimate of actual
volatility. When the yellow line drops below the blue line, VIX is
is underestimating actual volatility.
How to trade this? Maybe something like the following (in
pidgin Easy Language:
CORR = Pcorrelation(VIX[5] - VIX[10], 100*(OEX/OEX[5] - 1), 10);
if CORR > 0.0 then begin
if marketposition <= 0 and OEX > OEX[5] and VIX < VIX[5] then
buy ("B") on close
else if marketposition >= 0 and OEX < OEX[5] and VIX > VIX[5] then
sell ("S") on close;
end else begin
if marketposition > 0 then exitlong ("BX") on close;
if marketposition < 0 then exitshort ("SX") on close;
end;
I haven't tried this, but it looks like an interesting place to start.
--
| Gary Funck, Intrepid Technology, gary@xxxxxxxxxxxx, (650) 964-8135
X-Zm-Content-Name: vixstudy.gif
Content-Description: OEX v. VIX study
Content-Type: image/gif ; name="vixstudy.gif"
X-Zm-Decoding-Hint: mimencode -b -u
Attachment Converted: "c:\eudora\attach\vixstudy.gif"
|