|
PureBytes Links
Trading Reference Links
|
I couldn't see the relative weights on the url Gary posted. However, the
following
showed the highest unweighted correlation in descending order when run
against my stock database over the last year:
Weekly - MWD, CS, TIF, C, SANM, HWP, MER, MOT, MXIM, XLNX, LEH, LLTC, TNB,
SV, HDI, NXTL, CR, CTXS, JDSU, ADI, NOVL, TEK, BSC, SCH, SEBL, YHOO, KRB
Daily - MOLX, SCH, MWD, C, LEH, MER, GE, CSCO, BSC, JPM, ORCL, AXP, ADCT,
MXIM, HWP, CS, LLTC, CTXS, EMC, JDSU, SUNW, TROW, SEBL, TYC, BEN, CMVT,
XLNX.
----- Original Message -----
From: "Jack Griffin" <jack_2231@xxxxxxxxx>
To: "David Jennings" <DavidJennings@xxxxxxxxxxxxx>; "ed garib"
<egarone@xxxxxxxxx>
Cc: <omega-list@xxxxxxxxxx>
Sent: Wednesday, July 25, 2001 1:09 PM
Subject: Re: creating an index
>
> --- David Jennings <DavidJennings@xxxxxxxxxxxxx>
> wrote:
> > However, as a riule of thumb, I have found that the
> > 10 largest constuents
> > are sufficient to mimic performance..
>
> David or others in the know, do you think this applies
> to the popular indices as well? I mean if I were
> doing technical analysis of the S&P 500 and I had a
> hole in my data, would it be OK to fill it with an
> appropriately weighted average of the top 10 stocks in
> the S&P?
>
> Jack
>
>
>
>
>
|