RE: "Trading Ranges" - or how for SPC practitioners to make money from stock markets

["Alan Meekings" <alan@landmarkconsulting.co.uk>]

Stephen,

The problem with charting stock prices, as I'm sure you know, is that
the data is, technically-speaking, "highly auto-correlated" - that's to
say, today's stock price is highly dependent on yesterday's stock price.

A similar equivalent is the daily stock on shelves in stores or
warehouses.  Today's stock levels are obviously highly dependent on
yesterday's holdings.

This is very different from manufacturing or service processes where
each item of output is subject to common-cause and special-cause
variation.

If you've got a lot of time, Steven, together with an on-line source of
stock prices and a very powerful computer, you might like to explore the
correlation between movements in stock price that register as special
events not on the XmR chart of individual values but on the moving range
chart, and then relate these special events to subsequent "corrections"
for the over-reaction of markets.

Just for fun, Brian Wood and I followed the price of one of the UK's
leading supermarket stores back in 1995, and reached the conclusion that
we could have made a boat-load of money by using XmR charting (or what
we would call "iCharts") to predict the over-reaction of stock markets.

Sadly, we've both since been too busy to test if our observations for
one stock price in the UK would translate to markets in general.

However, if you do have success with this approach, Steven, you may wish
to recognise us in your Will with a small mention and a very big
donation to charity!!

Regards,

Alan
 
Mobile: +44 07785 258 741
Email: alan@landmarkconsulting.co.uk
 
-----Original Message-----
From: clauson@deming.ces.clemson.edu
[mailto:clauson@deming.ces.clemson.edu] On Behalf Of sbyers@wirb.com
Sent: Monday, April 21, 2003 11:45 PM
To: den.list@deming.ces.clemson.edu
Subject: "Trading Ranges"

Today's (4/21/03) WSJ has, on page C1, a graph of the Dow
Jones performance for some months. The accompanying article
is about Trading Ranges. When I look at the chart, it seems
ready to be a control chart, yet there is no effort at
statistical analysis in the article or on the chart. Am I 
missing something? 

I think a see a relatively stable process. The second head-
line says "Stocks Become Mired Within Narrow Band and Can't
Get Out." How about that? Further, "a trading range is
a band in which stocks sometimes get stuck, and from which
they have trouble breaking out. Investors who learned about
stocks in the 1990s didn't meet up with many trading ranges;
stocks rarely got stuck and mostly went up." Now, is that
the kind of helpful analysis one would expect from the WSJ?
It really sounds like the shoe manufacturer we all know
about, who replied to an inquiry about variation in the
number of defects, "Some days are better than others."

In the article, some bullish analysts warn "that, [even] if
stocks put in sharp gains now, they may have trouble 
hanging on to them."

As written, the article is an uninformative drive through
all too familiar terrain while gazing in the rear view
mirror. I think it could have been much more interesting.
Maybe it will show up on Tufte's web page as one of his 
"graphs of the day." And I hope Ken Macur might comment.

Regards,

Steven Byers
Olympia, WA

Message posting through the Clemson CQI Web Server.