Hello John, I don't use Worden so I can't run a test but theoretically I think you can do it.
AB can be treated as any database and it is robust, especially for EOD databases (mixing EOD and intraday is not such a great idea IMO). Often experiment is the best answer.... as long as you backup your good database and play around with a spare. AmiQuote/Yahoo Yes the history is deep ...it varies with symbol ... I have 10 years ... some went a lot further. The problem might be matching the data type. Worden has subjectively adjusted data? Yahoo has either div and event adjusted OR raw data (your choice). Yahoo does contain some errors so it will not deliver pin point accuracy for algorithmic trading e.g. I have seen a 50 cent div entered into Yahoo as 50 dollars. Yahoo Raw is more likely to be accurate since Yahoo do the adjusting themselves while the raw data comes straight from CSI. IF you are going to use Yahoo then you do need to learn the tricks of the trade (price adj and volume adj). There might be more than one way to do it. Try this to start. Export your Valueline index ASCII file. Start a new local EOD database. ASCII import your Valueline index to the new database. Get a list of Yahoo symbols for the stocks you are interested in, enter them to your new database and AmiQuote/Yahoo data for them. Pretty certain you could also append Yahoo data to a Worden sourced AB database as long as it is EOD ... the only prob would be any symbols that don't correspond, which is usually only the indexes. brian_z --- In [email protected], "john10987654321" <john10987654...@...> wrote: > > I am using Amibroker with the Telechart plugin. Recently Worden > stopped including the Value Line Geometric Index symbol in its list of > items. I want to backtest a strategy based on the performance of this > index. Is there a way to import historical prices going back, say, 15 > years, into Amibroker from Yahoo while still using the Telechart > plugin for all my stock symbols. I see that Amiquote can be used for > EOD data, but is that just for the most recent past or just today's > data? Any suggestions much appreciated. >
