|
Dan,
Yes, for the purpose of comparing results, since
protfolios are generally not representative of the S&P weighting, the
comparison would be more meaningful if one compared portfolio resuls to a not
non cap weighted average ... unless of course one purposefully tries to match
the S&P.
For simplicity's sake however, it may be just
creating extra work. As long as you are aware, for example, that if your
portfolio return is say 12% and S&P has 10%, your returns may still be
subpar (at least in today's environment). When the large caps start doing better
than rest of market, situation will be reversed.
----- Original Message -----
Sent: Wednesday, July 20, 2005 7:49
AM
Subject: RE: [amibroker] Re: Tomasz,
GetExtraData() Array Question
Ara,
Would it be more
reasonable to use an S&P 500 equal-weighted index? I.e., either one
provided by the data vendor such as “SPEW-X” in TC2005 or a composite created
in AB?
Regards,
Dan.
Just a quick comment on comparing test / investmet
results to S&P500.
Since S&P is capitalization weighted, it will differ
from any portfolio
results as
those would not be capitalization weighted.
If one took the average return of all S&P components
equally weighted, the
results
would be considerably higher because of the large cap
lower
performance.
----- Original Message -----
From: "sgfuchs"
<sgfuchs@xxxxxxxxxxxxx>
To:
<amibroker@xxxxxxxxxxxxxxx>
Sent: Wednesday, July 20, 2005 2:40
AM
Subject: [amibroker] Re:
Tomasz, GetExtraData() Array Question
> RJS
>
> AB
provides a list of QP datafields which can be accessed
via
> getextradata function -
see helpfile "Getextradata" function.
>
> In
order to retrieve "historical data" for these fields (
if
>
provided
> by QP )you need to
use also the "ref(Getextradata("XXXXX")" function
> and to enter the days/periods for how long AB should
look back in
> the QP
database.
>
>
EPSRank i.e. is available back to January 8th. 02
>
> I
have created a fairly complex strategy / Exploration
/
> Tradingsystem around the
datafields available from QP which I can
> backtest in AB because there's no other way to check
the validity of
> stock
selections performed in SPP based on fundamentals. However,
I
> needed to create a lot own
"datafields" and ratios to be calculated
> by AB in order to "simulate" rthe selection and
ranking process in
> SPP/ HGSI.
It seems, that this system works quite well in real
live
> trading. But the
possibilities for backtesting are still very much
> limited due to the fact, that I cannot access certain
QP datafields
> in AB ( in
particular the ranking values and A/D letter ranks )
>
> Hence
I would second the proposal to include some more QP
> datafields - at least those, which one can already
access also in
> Excel via the
QP Excel plugin. For most other fields, I fear, they
> are calculated during each update and / or no
historical figures are
>
kept.
>
>
That's one of the drawbacks of QP / HGSI - backtesting
of
>
fundamental
> strategies or
combined fundamental / technical strategies is
> impossible without 3rd. party software like AB - and
even this
> combination is
rather limited.
>
> I
used Vectorvest Online including their ProTrader application for
a
> 5
> week trial because it allows a quite convenient
backtesting of
> fundamental
strategies because it stores historic fundamental
values
> and their ratings,
partially back to 9 years. The software is very
> easy to use and very stable. However, TA
is
> However - 60.- etxra a
month + a 495.- USD one time fee for
> Pro-Trader
> in addition to my current data subscriptions is a bit
too much for
>
me.
> ( I use already AB,
Omnitrader, QP and TC2005 )
>
>
123portfolio.com allows also portfolio level backtesting on
more
>
than
> 500 fundamental andf
technical parameters, including a quite
> sophisticated ranking system for selection of trades.
Several users
> over there are
very educated and experienced. Some of them have
> already posted their real live portfolio results vs.
hteir
>
backtesting
> strategies. At
least from these documents, one can draw the
> conclusion, that this stuff works - and it has to,
because it makes
>
no
> sense to have a portfolio
turnover of 500% a year with mediocre
> results.
>
> For
what I can see, 123portfolio standard folio's outperform the
S+P
> by about 40% a year. Take
this with a grain of salt - almost all
> strategies are small cap related and hence the SP 600
or RUT would
>
be
> a more realistic
comparison for perfomance.
> I
compare those results with either small-cap mutual funds or
the
>
new
> Powershares ETF's ( i.e.
PWO, PWC ) because they offer a mechanical
> way to invest in a quantitative strategy without the
hassel of a
>
500%
> portfolio turnover /
year and no tax problems due to shortterm
> gains.
> Up 'til now, their performance is very impressive for
Indexbased
>
ETF's
> ( Intellidex Indexes )
see www.powershares.com
>
>
regards
>
Stefan
>
>
>
> ---
In amibroker@xxxxxxxxxxxxxxx,
"razorruss17" <rsebring@xxxx>
> wrote:
> > Please Help!
> >
>
>
> > Gary over at Quotes-Plus
told me on QP's message board that QP
> > database contains historical earnings and
revenues data fields. If
> >
this is the case shouldn't we be able to use the
GetExtraData()
> > command
to extract array's beyond the QRS, and EPSRank
arrays?
>
There
> > seems to be a
strong interest in many QP and Amibroker users to be
> > able to accomplish this. What's the verdict,
Tomasz? Anyone??
>
>
>
>
> > RJS
(Russ)
>
>
>
>
>
>
Please note that this group is for discussion between users
only.
>
> To
get support from AmiBroker please send an e-mail directly
to
> SUPPORT {at}
amibroker.com
>
> For
other support material please check also:
> http://www.amibroker.com/support.html
>
>
>
Yahoo! Groups Links
>
>
>
>
>
>
Please note that this group is for discussion between users only.
To get support from AmiBroker please send an e-mail directly to
SUPPORT {at} amibroker.com
For other support material please check also:
http://www.amibroker.com/support.html
YAHOO! GROUPS LINKS
|