|
PureBytes Links
Trading Reference Links
|
b & wd,
Since I don't use breadth data, I have to acknowledge some level of
ignorance relative to breadth measurements.
I looked at the following link:
http://www.tickquest.com/NeoBreadth/marketbreadth.html
and had to wonder why AB could not be used with the Addtocomposite
feature to create your own breadth measurements? The only limiting
factor might be volume calculations which would overflow the AB
fields, but which could be cut with a divisor to get them to fit.
So I would ask, what specific breadth measurements are problematic
with the data suppliers?
Regards,
Dennis
--- In amibroker@xxxxxxxxxxxxxxx, b <b519b@xxxx> wrote:
> wd,
>
> Thanks for the heads up about certain aspects of CSI's data
> as it deals with breadth, etc. I have filed your post in my
> CSI notes so that if I consider developing a breadth
> system, I will know what to watch out for.
>
> WHY CSI DATA IS ESSENTIAL FOR ME:
>
> My preference for CSI stock data arises from items that are
> essential to the type of system development I focus on.
>
> CSI has actual as well as adjusted prices and because some
> of my filters are price sensitive I need to know the actual
> trading prices of stocks for back testing.
>
> Secondly CSI has inactive as well as active stocks, and I
> need these as well because my systems give biased back test
> results if I only look at those stocks that have survived.
>
> CSI's CLEAN AND UNCLEAN DATA:
>
> That does not mean that CSI is perfect. From what I have
> seen, they appear to try as hard as any other provider to
> give accurate data. Your pile of pens attests to their
> desire to correct any mistakes that come to their
> attention. That pile of pens also attests to the fact that
> there are mistakes.
>
> As a CSI subscriber I have noticed that when I do a daily
> update, CSI sends the previous day's data to me as well as
> today's. To my mind this implies two very important things
> - first, CSI has a passion to give users the best data
> available AT THE TIME, and second, that there MAY BE
> mistakes in the daily data when it is first delivered to
> users.
>
> The possibility that CSI's data for the current day may
> have some mistakes that will not be corrected until
> tomorrow (or even later) has implications for my back
> testing and for my actual trading.
>
> TRADING IMPLICATIONS:
>
> I might get a buy signal for a stock based on bad data in
> today's down load for that stock. I may then buy the stock
> at tomorrow's open and then discover that there was no buy
> signal when I down load tomorrow's data which includes
> corrections for the previous day's data.
>
> What to do? If I knew of a cleaner data provider I would
> switch, but I do not. So what to do with a buy signal
> knowing that the data might not be perfectly clean?
>
> I build the possibility of being wrong into my money
> management plan.
>
> BACKTESTING IMPLICATION:
>
> When I run back tests, I know that I am getting somewhat
> artificial results since I am using "cleaner" data in
> testing than I will have when I actually trade. I see no
> solution for this, so I just take back test results with an
> extra grain of salt.
>
> Ideally, I would like CSI to provide clean data (which I
> would use for profit and loss calculations) and "raw" or
> uncleaned, never corrected data, which I would use for
> signal generation. That way my back testing would be closer
> to the conditions I have to operate on when actually
> trading.
>
> CONCLUSION:
>
> CSI is not perfect. But, as far as I know, it is still
> better than the alternatives "most of the time" for "most
> things" for the type of trading and system development that
> I do.
>
> However, I am thankful for users who point out limitations
> in CSI data (ie, the breadth problem noted by wd).
>
> b
>
> --- wd <wdjd@xxxx> wrote:
> >
> >
> >
> > > Maybe - BUT how will you know which signals are valid?
> > Oh yeah - you'll
> > > know after they fix it (if they do) and your signal
> > disappears or your
> > trade
> > > tanks and you spend countless hours trying to figure
> > out why only to
> > > discover a bad quote. But its only money, eh?
> >
> > I use quite a bit of breadth data in my system. I was a
> > QP user for about 4
> > years. Before then I used TC2k, AIQ and a host others. I
> > did lose money one
> > time (that I know of) because of bad breadth data when I
> > was using QP. I
> > complained several times on the QP list and other members
> > checked into my
> > complaints and found them valid. QP beat around the bush
> > for several months
> > while this was going on and I switched to CSI (I was told
> > later by another
> > member that QP did finally clean up their breadth data).
> >
> > I switched to CSI because of their reputation for clean
> > data. I found most
> > of their data to be clean (so far as I could tell), but I
> > did find quite a
> > bit of bad data in their breadth and indices. Not just
> > minor indices, but
> > even the SPX.
> > CSI offers a CSI pen if you make them aware of any
> > discrepancies. I have a
> > desk full of them.
> >
> > I now use a paid data service for my breadth and indices
> > that is far cheaper
> > then the above two and I get my stocks/ funds from the
> > various free data
> > sources. I rarely have data issues with the free sources
> > and when I do, I
> > check with a couple of the other free data sources. This
> > may happen one or
> > twice a year.
> >
> > I am NOT saying this is the way to go for anyone, but it
> > is most definitely
> > for me. I don't have *full* confidence in my present data
> > sources, but I do
> > have more confidence in them then I do with some of the
> > paid sources.
> >
> > Jim
> >
> >
>
>
> __________________________________
> Do you Yahoo!?
> Yahoo! Search - Find what you�re looking for faster
> http://search.yahoo.com
Send BUG REPORTS to bugs@xxxxxxxxxxxxx
Send SUGGESTIONS to suggest@xxxxxxxxxxxxx
-----------------------------------------
Post AmiQuote-related messages ONLY to: amiquote@xxxxxxxxxxxxxxx
(Web page: http://groups.yahoo.com/group/amiquote/messages/)
--------------------------------------------
Check group FAQ at: http://groups.yahoo.com/group/amibroker/files/groupfaq.html
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:
http://docs.yahoo.com/info/terms/
|