|
PureBytes Links
Trading Reference Links
|
Walter,
>Can someone please explain the mind set of setting up a stock
>database?
The US-Stocks database is a basic database suitable for learning AB
and in some cases trading (some traders use it and some don't, or
won't).
As a basic database it doesn't allow users to apply some of the finer
nuances of trading e.g. backtesting delisted stocks etc. Accounting
for the nuances comes with more experience and requires bought data.
Initially for basic backtesting databases:
(I stress this is a personal opinion only - others have different
ideas).
1) The main problem is getting long enough histories to work with -
you need to keep half your data for out of sample testing so you can
only experiment with the first half of your data.
So, as a rule of thumb, the more data you can download the better - I
have no idea how far back Yahoo goes - it varies from symbol to
symbol.
The caveat there is that the nature of the markets might have changed
so some traders don't like to use data pre 1990's (assuming that
widespread computer analysis kicked in around then).
2) Not all symbols will have deep histories and some will be new
listings, or have undergone name changes, so what you get will vary.
I recommend:
1) that you take a good long history (I only took the most recent 10
years for the US-Stocks database but I supplement it with other
markets and other data - I also go about my evaluations differently
to most others so I don't use as much data - 6 hours download on slow
broadband and 420MB on disc) - you could take 20 years if you want or
2 * 10 years.
2) download it in one hit and don't add to it again (keep it backed-
up and quarantined) - don't use daily updates as they can mess up the
data.
3) only use the oldest 10 years of data while you are learning and
keep the most recent 10 years until you have more experience with
backtesting - especially system design and evaluation.
If you use up all of the most recent data know you will regret it
later.
4) start a shorter separate database that you update daily for
trading (or paper trading etc) i.e. use it for following the markets
and not backtesting - dump it every so often and start again - you
don't need much of a history for scanning and following your trades
etc - say you are using a 250MA indicator then you only need 250 bars
plus a bit.
Use the current data mode to download the daily bar in your live
trading database - you will get the close price earlier than if you
wait for the historical server.
There is more to it than that but you can't absorb it all in one step.
> Or would I be better off only downloading 2 years of EOD data for
>each
> symbol and then if I need more data for a specific symbol then I
>would
> simply go back and download it from Yahoo.
No, don't play around with backtesting databases - get em and keep em
clean and use them for backtesting only - good data is precious.
> when I brought up the chart, I noticed between the years of 1928 and
> 1930 there are dashes (something like open or close markers without
> the bars).
I'm not certain but bars like that are used for data padding - say
the market didn't trade on that day or the data wasn't recorded for
some reason then data providers sometimes fill the gap with an O-C
bar, especially if it is the same as the O or C from the day before.
I don't know why two years would be like that - some one else might
know.
brian_z
--- In amibroker@xxxxxxxxxxxxxxx, "Walter Lepore"
<electricwally77@xxx> wrote:
>
> Hi Members
>
> Couple of quick questions please.
>
> These questions refer to End Of Day Data please.
>
> 1) I am setting up my EOD stock database for the first time. I am
> given the opportunity to download 7700 stocks into my US-Stocks
> database using AmiQuote as a downloader from Yahoo historical data.
>
> The AmiQuote downloader will download EOD data on any time frame I
> input. Ideally it would be nice to download all available data from
> all stocks and indices. However not only would it take a couple of
> days but would probably take up a ton of space as well. I have the
> time and the space for this procedure but is this the correct way to
> start building a data base?
>
> Or would I be better off only downloading 2 years of EOD data for
each
> symbol and then if I need more data for a specific symbol then I
would
> simply go back and download it from Yahoo.
>
> I understand that the time frame I should download is dependent on
> what i am trying to back test.
>
> Can someone please explain the mind set of setting up a stock
> database? In other words, its the time frame for each symbol I'm
> deciding on.
>
> 2) I downloaded historical EOD data on The Dow Jones Industrial Avg.
> ^DJI. The data goes back to 10/1/1928. The download was successful
but
> when I brought up the chart, I noticed between the years of 1928 and
> 1930 there are dashes (something like open or close markers without
> the bars). From 1930 onwards to Dec. 2007 the chart is normal(shows
> OHLC Bars). Why are these dashes present instead of OHLC Bars?
>
> Thank You
> Walter
>
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 NEW RELEASE ANNOUNCEMENTS and other news always check DEVLOG:
http://www.amibroker.com/devlog/
For other support material please check also:
http://www.amibroker.com/support.html
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/amibroker/
<*> Your email settings:
Individual Email | Traditional
<*> To change settings online go to:
http://groups.yahoo.com/group/amibroker/join
(Yahoo! ID required)
<*> To change settings via email:
mailto:amibroker-digest@xxxxxxxxxxxxxxx
mailto:amibroker-fullfeatured@xxxxxxxxxxxxxxx
<*> 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/
|