Richard,
<<"he is after the price of a stock AS IT WAS TRADING,
say 60 days ago, but NOT reflecting the effects of any
subsequent splits."
Sounds like he is. I am trying to figure how why? >>
I have find a modest but real correlation between the price level
of a stock and its intermediate future price performance.
For example, say I look at the 4000 or so common stocks
in the qp2 system which have 2 years of history and sort
them by price as of 6 months ago, into 100 groups of 40
where group 1 is the 40 stocks with the highest prices,
and group two is the next 40 highest on price and so on down
to the 40 lowest on price.
If I then calculate summaries for each of these 100 groups
which include average 6 month forward investment return,
as well as average market cap weighted 6 months forward
investment return, these investment return vectors for the
100 such groups will show highest returns for the highest
priced groups, and declining returns for the intermediate priced
groups, and lowest returns for the lowest priced groups.
There may irregularities in how these %return vectors
decline from high price groups to low price groups,
but they are relatively minor. There are far fewer such
irregularities when the returns are taken cumulatively thru the
100 groups.
The positive bias for selecting a $90 stock (for intermediate
or long holding periods) may be small, but these biases exist
for many variables and taking a fair number of them (20 or more)
has resulted in results meaningfully better than the sp500
over the last 13 months.
These biases (investment advantages) also let me use stocks selected
for inferior performance to hedge the long position in order
to be relatively indifferent to the direction of the market.
<<Of course the concept is unrealistic, and will cause him
more pain than gain.>>
I have not found the concept unrealistic. I grant that my
approach has involved a large analysis effort on my part,
but this is in large part pleasurable and the potential reward
justifies the effort for me. |
