From: John Conover <john@email.johncon.com>
Subject: Re: Optimal Stock Market Strategy Re: SGP versus Switching Campaigns;
wisest strategy
Date: 3 May 1999 15:21:29 -0000
Pertti Lounesto writes:
> Archimedes Plutonium wrote:
> >
> > 1) Pick the two best industries, make one biology and one physics, for
> > contrast
> > 2) Pick the best company in those two industries
> > 3) Pick the large capitalization, high dividend and rather depressed
> > 4) Operate a _switching campaign_ on these few companies in the
> > portfolio
> > 5) Vary the allocation of the assets through the years, some years 100%
> > in one, other times 50-50 and in-between
> > 6) Change these best companies in the two industries, when you think
> > another better company in that industry emerges
>
> With two companies only you still have a high company risk.
> I would say five companies is minimum, if you want to reduce
> the company risk close to zero (so that only market risk
> remains). On the other hand, one person does not have enough
> resources to become acquainted with more than ten companies.
> Thus, pick up 5 to 10 companies, each from different industries,
> the leading brand from each industry. After this choice you
> have two alternatives: 1. buy and keep, 2. active trading, maybe
> even daytrading. In the alternative 2. your portfolio requires
> a lot of effort, maybe fulltime work. I would recommend the
> alternative 2. only after you have a few years of experience
> about the stock markets. As for the choice of industries, I
> would recommend: internet, electronics, dietary supplements,
> and pharmaseutics.
>
Yes, as discussed earlier, 10 is about right. Under the presumptions
of the EMH, (random walk, and market efficiency,) volatility equates
to risk, and the volatility of a portfolio is the root mean square of
the volatilities of the individual stocks in the portfolio. So the
portfolio's risk would be decreased by a factor of 1 / sqrt (2) to
1 / sqrt (10).
Under the presumptions of the EMH, a portfolio's value can grow
faster, for a while in the short term, with only 2, but given enough
time, the statistics of the higher risk will catch up, and, sooner or
later, the portfolio's long term performance will be worse than with
10.
There is a caveat, though. The scenario is more complicated than
that. What one really wants is the average of the long term marginal
returns to be the square, (again, presuming an EMH model,) of the root
mean square of the marginal returns. Unfortunately, most of the stocks
in the world have the average > root mean square, (showing investor
risk aversion,) and optimality can not be obtained. So, lowering
portfolio risk is the next best alternative.
There is a graph of it at:
http://www.johncon.com/john/correspondence/980814222509.25795.html
and long term simulations on the historical ticker of the US exchanges
seem to confirm the EMH reasonably well, (depending on who is telling
the story, of course-I personally don't like the paradigms of the EMH,
but many do.)
John
BTW, in the graph, the portfolio's gain in value is twice the gain in
value of any stock in the portfolio. On average, a stock in the US
exchanges has 51 up movements for every 49 down movements. If one
thinks about it, having a portfolio of many stocks will exploit such
combinatorics. The gain in value of the portfolio will be larger than
the gain in value of any stock, and the risk to the portfolio will be
less than the risk of all stocks. Kind of a maximizing gain while
minimizing risk technique at the same time. Peter Lynch was probably
right when he said that making money on Wall Street is easy-keeping it
is the hard part, (meaning that managing risk is how one maximizes
gain.)
--
John Conover, john@email.johncon.com, http://www.johncon.com/