Message Font: Serif | Sans-Serif
 
No. of Recommendations: 20


"HEY GrandpaRalph" yelled GrandmaSue from the Family Room.

"Whatever happenned to that market-beating strategy soundly based
on statistical analysis you've been working on for all these months?"


It has been about 6 weeks since Part 5, in which I tied all those
individual "quality indicators" together into an overall
quality/valuation function, I thought to myself.

That was back in post: http://boards.fool.com/Message.asp?mid=14886700

"Maybe I should share the latest model refinements with the few Fools
left who actually talk about stocks and strategies for picking them"
,
I yelled back.

I thought about the last few months and realized that for some reason
I haven't posted much lately. But it's hard to get motivated when it
seems folks are more interested in bashing either TMF, or the fallen
tech giants, or each other lately. Doesn't take long to get fed up
with the MSFT and similar boards under these conditions. And poor
Orangeblood thought he could control that unruly crowd?!. Oh well,
probably just tension due to current market trends. Still, I wish
some of my favorites, like Buckwheat3 and his little furry Cisco would
come back. I sure hope they didn't die in the desert. Or take up
hermithood.

"Maybe I should give the model a report card, too", I continued.

"I know 6 weeks isn't long, but for crying out loud, this is
statistics picking stocks....well, a little additional Due Diligence
is done, like looking for integrity issues in the management team and
nasty lawsuits, but it should start working right away, right?...
RIGHT?"

"SUE!! ARE YOU LISTENING?"


No, she's watching that dumb "Friends" show again, I realize.

"OK, I'll try to put together an update and a report card...
thanks for the interest."


*********************************************************************
*********************************************************************

Hi again RuleMaker and Valuation Fools...

Since Part 5, I have continued to refine the model, add more
companies, throw out companies of "low" quality, (since a low quality
company at ANY price is still too expensive) and have begun to
track various portfolios that the model said were either highly
overvalued or highly undervalued.

MODEL REFINEMENTS:

1) What to look at:

One of my goals was to understand what parameters were of "value" in
the market, and which ones were not, so that I could stop tracking
stuff that doesn't really seem to influence valuation. At this time,
the parameters that affect valuation to the greatest extent would
appear to be (in order) Net Margin, CashKing Margin, (Operating Cash
Margin is almost equal, but I don't think both should be used
simultaneously) Gross Margin, Future Growth rate projections,
Cash/Debt Ratio, and TTM revenue growth rate.

2) Better defining the individual quality relationships:

In the past, I only considered linear relationships. But since the
statistics package I bought will do polynomial regressions, I figured
hey, why not live on the edge, and plot a few best fit curves. Can
you imagine the shriek of glee I let loose when I first looked at
Price/Sales Ratio vs. Net Margin when plotted as a second order
polynomial?

http://people.mw.mediaone.net/ryoungbl/nm061601.gif

(Note: EBAY was omitted from above graph because its P/S ratio of
36.3 "broke the curve". EBAY is however in the database.)

BUILDING MOCK PORTFOLIOS:

Approximately monthly, I make an intense effort to update the data in
my spreadsheets with all available data, (whether in a 10Q or just an
earnings release) and then take a "snapshot" of all prices, growth
predictions, and other parameters. I then rerun all the regression
analyses for each parameter, construct a new quality function that
reflects the current market, compare each company's current valuation
to the model's "ideal" valuation, and then make an ordered list of all
the companies, from "Most Undervalued" to "Most Overvalued". I then
take an imaginary $20,000 and invest half of it in the 10 most
undervalued, and half of it in the 10 most overvalued. Here are the
6 portfolios I've built and am tracking:

04/28/01 05/24/01 06/16/01
UnderValued OverValued UnderValued OverValued Undervalued OverValued

AMAT AMGN AMD AMGN BARZ AMGN
BARZ AOL BARZ AOL CA AOL
BRCM BRCD CAKE BRCD COH BRCD
CSCO CIEN CHRT CREE IDTI CREE
IDTI CMVT COH EBAY JDSU EBAY
LSCC EBAY ELY LLTC LSCC KKD
MACR MERQ IDTI MERQ MACR MDT
PHG MSFT LSCC MSFT NTAP MERQ
RATL PAYX MACR PAYX RATL PAYX
SYMC QCOM OSI QCOM SYMC QCOM

NOTE: The 6/16/01 portfolios are the first to use non-linear
relationships in the quality functions. I was able to improve the
model's correlation to the real market by about 4 or 5 percent, and
noticed that while virtually no change occurred on the "undervalued"
side of the list, there were some differences in the companies
identified as "overvalued" by this method.

REPORT CARD:

OK, let's look at how each of these portfolios is performing vs. the
major indices.

First, the 04/28/01 portfolios:

Value on 04/28/01 Value on 06/16/01 Percent Change

UnderValued4/28 $10,000 $9,149 -08.51%
OverValued04/28 $10,000 $10,135 +01.35%
Dow 10,810 10,623 -01.73%
Nasdaq 2,076 2,028 -02.31%
S&P500 1,253 1,214 -03.11%

...uh-oh...

My group of "undervalued" stocks, which I had hoped would get
"discovered" by the market, got trounced by all the major indices.
To add insult to injury, my "overvalued" stocks not only did better
than the indices, but actually went up as a group in a down market.
Maybe I should re-read RuleMakers, RuleBreakers or something. I
vaguely recall that when a stock is called overvalued... well, never
mind, I'm not an analyst anyway.

Must be because this was my first attempt, and a lot of refinements
were made after 4/28.

So, on to the 05/24/01 portfolios:

Value on 05/24/01 Value on 06/16/01 Percent Change

UnderValued5/24 $10,000 $8,694 -13.06%
OverValued05/24 $10,000 $8,700 -13.00%
Dow 11,122 10,623 -04.49%
Nasdaq 2,282 2,028 -11.13%
S&P500 1,293 1,214 -06.11%

...uh-oh...

My "undervalued" stocks got just barely edged out in performance by
the "overvalued" stocks. And BOTH portfolios, representing two
opposite ends of a GREAT valuation model, got clobbered compared
to the indices. Is there no hope for this process?

But wait, I've introduced those polynomials into the 6/16 portfolio,
and at this writing, I've already got one day's worth of data. I'll
go check the TMF portfolio pages...

Let's see...

UnderValue06/16/01: Down 4.18% on 06/18/01 :-(
OverValued06/16/01: Down 0.71%
Dow: Up 0.2%
Nasdaq: Down 1.96%
S&P500: Down 0.49%

...uh-oh...

The undervalued stocks got killed, the overvalued stocks held up
reasonably well vs.the 3 indices.

*********************************************************************
*********************************************************************

"Hey, Sue!"

"I've got that Part 6 post finished, and once again, I've shown what a
market GENIUS I am. You won't believe this new shorting strategy I've
developed!"


Ralph





Print the post  

Announcements

What was Your Dumbest Investment?
Share it with us -- and learn from others' stories of flubs.
When Life Gives You Lemons
We all have had hardships and made poor decisions. The important thing is how we respond and grow. Read the story of a Fool who started from nothing, and looks to gain everything.
Contact Us
Contact Customer Service and other Fool departments here.
Work for Fools?
Winner of the Washingtonian great places to work, and Glassdoor #1 Company to Work For 2015! Have access to all of TMF's online and email products for FREE, and be paid for your contributions to TMF! Click the link and start your Fool career.