(followups: I removed the SAS newsgroup, and added the
sci.stat.consult group which had received a separate copy of the
question.)
On Sat, 13 May 2000 06:59:45 -0000, "John Anderson"
<[EMAIL PROTECTED]> wrote:
> I work for a bank. There are 2 groups of accounts chosen randomly when they
> were booked. Those 2 groups of accounts are comprised about 20% of the
> whole portfolio. One group is called the champion and the other is called
> the challenger. The champion has the same strategies as the rest of the
> whole portfolio, except the challenger, which has different stratgies. We
> try to have the challenger outperformed the champiom for at least 6 months,
> so we can implement the same strategies to the rest of the portfolio, then
> create a new challenger. I want to perform a statistical test (maybe
> t-test)to conclude that the challenger has outperformed the champion overall
> in a 6 month period. Should I use their average performances such as
> profits or revenues each month for 6 months (n=6 for the challenger and n=6
> for the champion, each month for each group is a data point or (n)), or
> should I compare by using each individual account each month for 6 months
> (n=10,000 accounts for the challenger, n=10,000 accounts for the champion).
> I apologize if my question is too naive; I took only 1 statistics class.
> Any suggestion, idea for the test or forcasting is greatly appreciated.
Past performance of a strategy in the stock market does very poorly at
forecasting future performance. That is especially true, in
competition with other strategies that share the loss/benefit of a
single market trend.
Different companies have different strategies; there is a *very* rapid
"regression to the mean" , according to what I have read, from "best"
to "mediocre". (I suspect that the WORST may continue as pretty-bad;
else, that would prove that my TIAA-CREF is wasting a million dollars
each on some top-executive salaries.) I am saying, experience says
that the time-series is not stable enough to be predictable, and
taking more statistics classes is not going to help.
The two guys who won a Nobel in economics (for figuring out the price
of Discounting, I think) needed a billion-dollar bailout of their
company. Apparently the way to stay rich is to have the government
take up the slack.
(There's one traditional distinction between the parties in U.S.
politics -- "Big-money" regularly buys or rents the support of the
Democratic party, whereas the Republican party is a wholly owned
subsidiary.)
--
Rich Ulrich, [EMAIL PROTECTED]
http://www.pitt.edu/~wpilib/index.html
===========================================================================
This list is open to everyone. Occasionally, less thoughtful
people send inappropriate messages. Please DO NOT COMPLAIN TO
THE POSTMASTER about these messages because the postmaster has no
way of controlling them, and excessive complaints will result in
termination of the list.
For information about this list, including information about the
problem of inappropriate messages and information about how to
unsubscribe, please see the web page at
http://jse.stat.ncsu.edu/
===========================================================================
