--
<<We certainly differ on how we look at the market. or evaluate results. You cut the
market to 4000 or more than 1/2, using some criteria.>>
No. The 4000 are ALL the stocks in the QP database which
contain 2 years of data. I do exactly the same thing you do when
you do a scan against the QP2 database. I apply a filter of my
set of weighted variables against the QP2 stocks and retain the
best 4 or 5 (for long postions) and the worst 4 or 5 (for possible shorting).
<<The inclusion or exclusion of a
filter might have far more effect on the results than the price level.>>
I did not say that price level is the most important
variable relative to future performance, just that I DID
find it useful (see table below). And that I could see why
Jim could want to find it in qp2.
<<The time frame of 6
months should be taken at a number of points since you are measuring price not market
conditions or an exit/entry system. The relationship should exist between groups with
decending price groups having smaller % profit. Because if I can understand what you
said, your work says that the higher the price at entry, the more return. Something I
have never seen>>
--Stocks priced March 13, '98-- --Stocks priced Sept. 11, '98--
--%Return as of Sept. 14, '98-- --%Return as of March 12, '99--
(Sept 11 '98) (March 12 '99)
Stock Price Range 6 Month Return 6 Month Return Note: %returns are group averages
----------------- -------------- --------------
70 & up -8.8% +28.6%
40 to 70 -13.4% +22.8%
20 to 40 -19.1% +11.3%
10 to 20 -23.0% + 7.5%
SP500 -4.4% +31.4%
The correlation between price level, and subsequent 6 month
performance is clear to me in both the downtrending stock
market ending Sept 14, '98, and the strong market period
ending March 12, 1999.
I have found a number of other variables positively
correlated to performance. I have weighted them and
applied these variables to all the QP2 stocks in order
to produce a weighted 'Value' per stock.
When I sort these stocks by 'Value' and compute their
3 month, 6 month or 1 year subsequent %returns, I find
the top group of 5 to 10 stocks regularly well exceeds the SP500.
I also find the poorest 5 to 10 stocks regularly far under
performs the long group, usually with negative returns.
This makes the 'poorest' group a nice candidate for hedging
and to reduce volatility. This trend, which is also present
within the 'Value' extremes, for me lends validation to
the hypothesis that what I am finding is not
statistical aberration.
While I have been finding these relationships for only 13 months
of QP2 data, this is all the 'as then available' data I have,
and this has been through up and down market periods.
--
|
