Re: SGP versus Switching Campaigns; wisest strategy

From: John Conover <john@email.johncon.com>
Subject: Re: SGP versus Switching Campaigns; wisest strategy
Date: 29 Apr 1999 08:31:31 -0000


Archimedes Plutonium writes:
> In article <7fug8i$abc$1@dartvax.dartmouth.edu>
> Archimedes.Plutonium@dartmouth.edu (Archimedes Plutonium) writes:
>
> > Stellar Performers
> > advantages
> > 1) once decided, they take the least amount of time whereas Switching
> > takes a large amount of time
>
> For most investors, this is the best strategy to follow since most
> investors really do not want to spend the time it takes to research
> companies and make decisions. Thus, the best strategy for them is to
> try to pick a stellar performer and to stick with the company. In the
> past 25 years, General Electric, Glaxo, Schering Plough, and Merck have
> all been stellar performers. Thus, buy the stock and lock it up and
> away and add more shares of the company as the years go by. I am
> talking about 10, 20, 30 years investment. In this strategy, though, it
> is crucial in picking a stellar performer.
>
>   My future strategy is to do a hybrid of two strategies-- stellar
> performer and switching campaigns. I will have at least 2 companies.
>

Just as a note, your strategy is probably sound as a long term
investor. You could probably pick up a few percentage points per year
by increasing your asset allocation to 10 stocks, keeping an equal
fraction of investments in each company, (ie., about 10% of your
equity portfolio in each company.) The companies should be selected on
the criteria, as you say, on largest market share, (which kind of
implies the largest capitalization-unless it is an Internet company.)

Simulations of the US exchanges over the century, last half century,
last quarter century, last decade, and last 5 years tend to indicate
that such a scheme would have beat all mutual funds-in the long
run-and would have come vary close to beating the major indices, (at
times exceeding the indices, which is difficult to do on a sustained
basis of 15, or more, years.)

The key concept is to adjust asset allocations among the 10, or so,
stocks such that each maintains a 10% fraction of one's equity
portfolio, (if you think about it, it is a dynamic buy-low-sell-high
strategy-that is why one would sell out of the "winners" and put money
in the up-and-coming, instead of letting things ride trying to pull
the last nickel out of the market, which is, usually, a costly
decision 50% of the time.) Another way of looking at it is that one is
hedging a single stock, against the other nine, (see the BTW, below.)
Most folks that run this strategy adjust their asset allocation
monthly, (or maybe quarterly,) or when any one stock gets more than
+/- 2% out of line.

The programmed day traders will do modestly better-but after they send
their broker's kids to collage, it is pretty close to a wash, (not to
mention that half will do much worse.)

        John

BTW, in case you are curious as to why the scheme works, it is because
the volatilities, (ie., the root mean square of the fluctuations in a
stock's value,) in the portfolio add root mean square, (duh.) The root
mean square is the risk of investing in the stock-at least according
to EMH dogma. The growth in value of the stocks in the portfolio add
linearly. So, it is entirely possible, (and frequently
happens-actually, almost always,) that the portfolio growth is larger
than any of the stocks in the portfolio. The theoretical numbers are
that one's portfolio should contain 50 stocks. However, there is
little marginal value in holding more than 10, (it looks kind'a like a
square root function on the number of stocks, and flattens out after
10, or so.) The reason for choosing stocks that are from companies
with the largest market share is that, for long term investing,
staying power in the market is an issue-and staying power is
proportional to market share. If one searches the exchanges, (maybe
using a computer to auto'magic it,) one can take the 6,000 stocks, (or
so,) on the US exchanges and narrow it down to about a hundred-which
still leaves a lot of choice for personal preference and knowledge of
specific markets and companies. However, such schemes are no
substitute for diligence with one's money, and the edgar database at
the SEC is quite helpful-http://www.sec.gov/edgarhp.htm, (if one can
read a general ledger-if one can't, one shouldn't be investing in
equities.)

As another little curio BTW, based on the assumptions of the EMH,
(ie., market efficiency and a non-linear random walk model as a first
order approximation,) if one picks 10 stocks, all equal, one's
portfolio would grow at almost twice the compound percentage rate of
any of the individual stocks in the portfolio. (Assuming a probability
of an up movement on any given day of 51%, down 49%, a root mean
square of the fluctuations of 2%, and a growth of 1.0003 per
day-typical for stocks on the US exchanges, over the past 25 years.) A
way of thinking about it, (other than information-theoretic stuff,) is
the combinatorics of the way a stock looses money-ie., what's the
chances of all stocks having a loosing day? The chances of all stocks
having a winning day is significantly higher. So, one makes a little
money off the fluctuations, in addition to the long term growth,
while, at the same time, mitigating risk, substantially.

--

John Conover, john@email.johncon.com, http://www.johncon.com/


Copyright © 1999 John Conover, john@email.johncon.com. All Rights Reserved.
Last modified: Sun Nov 14 00:01:17 PST 1999 $Id: 990429013144.4489.html,v 1.0 2001/11/17 23:05:50 conover Exp $
Valid HTML 4.0!