|
PureBytes Links
Trading Reference Links
|
you can simply test what the problem could be. Do a
backtest over a limited period and compare the trades.
One thing that could have gone wrong is that you
pad with the same symbol (see AA, settings, Pad and align data).
But in general to find what the problem could be
you have to start comparing the trades you get for the different databases and
go from there,
rgds, Ed
<BLOCKQUOTE
>
----- Original Message -----
<DIV
>From:
<A title=danielwardadams@xxxxxxxxx
href="">danielwardadams
To: <A title=amibroker@xxxxxxxxxxxxxxx
href="">amibroker@xxxxxxxxxxxxxxx
Sent: Friday, July 23, 2004 5:39 PM
Subject: [amibroker] Re: Simple
question
Does anybody else have problems getting comparable results
doing daily backtests of the same stocks with the same strategy, period,
and settings using the Yahoo and eSignal databases?I just tried
another period (365 days) and although the number of trades was comparable
(407 versus 443), the annual returns aren't even close (212.97% versus
72.09%).Would somebody please shed some light on these
descrepancies.Dan--- In amibroker@xxxxxxxxxxxxxxx,
"danielwardadams" <danielwardadams@xxxx> wrote:> If I
backtest the same portfolio of (approx 50) stocks with exactly > the
same strategy and settings over the same period of time, > shouldn't I
expect fairly comparable results with both the Yahoo and > eSignal
EOD databases?> > I would expect them to be close (acknowledging
things like splits and > data errors might be handled differently),
but my results aren't even > close.> > One test from
August 1999 to date gave the following results.> >
Yahoo: Annual return = 66.71%, number of trades = 2287.>
eSignal: Annual return = 52.05%, number of trades = 3977.> >
Although the eSignal data costs $$$, frankly I think I trust the >
Yahoo data more. I've reported problems with the determinism of the >
eSignal data before and haven't gotten any satisfactory feedback. >
(Example: Not forcing backfill on all stocks to the same date even >
though the data does exist). > > Which data should I believe?
Or, maybe, are both sets of data okay > and it's a matter of the
backtest code not working properly? > Backtesting is the ONLY reason I
want to use Amibroker but I'm about > ready to give up on
it.> > I'd be interested in anybody else's experiences on either
data > reliability or backtest reliability/determinism.>
> DanCheck AmiBroker web page at:<A
href="">http://www.amibroker.com/Check
group FAQ at: <A
href="">http://groups.yahoo.com/group/amibroker/files/groupfaq.html
Check AmiBroker web page at:
http://www.amibroker.com/
Check group FAQ at: http://groups.yahoo.com/group/amibroker/files/groupfaq.html
Yahoo! Groups Sponsor
ADVERTISEMENT
Yahoo! Groups Links
To visit your group on the web, go to:http://groups.yahoo.com/group/amibroker/
To unsubscribe from this group, send an email to:amibroker-unsubscribe@xxxxxxxxxxxxxxx
Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.
|