Historical Data

Discussions about the testing and simulation of mechanical trading systems using historical data and other methods. Trading Blox Customers should post Trading Blox specific questions in the Customer Support forum.
weatherman
Contributing Member
Contributing Member
Posts: 9
Joined: Thu Oct 16, 2003 8:09 am

Post by weatherman » Tue Jan 13, 2004 7:19 am

Thanks Moodaeng, have you used it to download data into veritrader or a similar format? I tend to use it for charts and live data. The historical data for equities is quite substantial as I’m sure it is for commodities and FX. It’s just a matter of pulling it out.

Moodaeng
Senior Member
Senior Member
Posts: 34
Joined: Thu Apr 17, 2003 4:23 am

Post by Moodaeng » Tue Jan 13, 2004 8:13 am

I retrieve the data directly in excel with the functions BLPH for historical and BLP for real time (Bloomberg calls these Worksheet Functions). I do my analysis in excel, so I don't have to export the data.

weatherman
Contributing Member
Contributing Member
Posts: 9
Joined: Thu Oct 16, 2003 8:09 am

Post by weatherman » Sun Jan 18, 2004 10:57 pm

Thanks Moodaeng. I’m set up on Bloomberg now. By the way what type of analysis do you do in excel.. if you don’t mind the question.

forex_kid
Senior Member
Senior Member
Posts: 47
Joined: Thu Apr 17, 2003 5:30 pm
Location: Sacramento, CA

Post by forex_kid » Mon Jan 19, 2004 12:37 am

Any Olsen Data users out there?

http://www.olsendata.com/index.html?standard_datasets

Looking at their page called "Forex Interval Data Pricing" under the "Historical Data" section leaves me with many questions about what people thing would be the most cost efficient parameters to choose for ordering intraday FX data.

Any thoughts out there on how granular you *really* need to get? I suspect it is probably worth it in the long run to get the full "6 field" option.

-Morgan

TradingCoach
Roundtable Knight
Roundtable Knight
Posts: 176
Joined: Thu Apr 17, 2003 9:52 am
Location: Sacramento, CA
Contact:

I use a NT service I built to gather eod futures data

Post by TradingCoach » Fri Jan 30, 2004 9:19 pm

email PM me if interested...

kalitka
Full Member
Full Member
Posts: 14
Joined: Tue Nov 04, 2003 4:13 am

Index Futures End-of-Day Data

Post by kalitka » Sat Feb 21, 2004 6:05 pm

I came up with a promising system for swing trading index futures. I have backtested the system based on indexes downloaded from Yahoo, and it shows quite dramatic results in particular on ^SPX. However, in the upper left corner of Yahoo Finance Page the link to S&P500 points to something called ^GSPC (what exactly is it ? who computes this index?). My system shows good results there as well but they differ significantly with ^SPX, with returns being inferior. Looks like ^GSPC accounts for gaps as opposed to ^SPX quoting so as Close always equal the next day Open which of course would distort results in real futures trading.
I also have actual dayly SP contract data and tested the system on them as well, but my data is not continuous, and it's much easier to test on indexes. However, my concerns are intraday futures volatilty, that index data my not reflect. Question: Is historical backtesting on indexes representative for actual index futures trading ? Dividends (reflected in indexe prices) are also a concern. Anyone can point me to the right direction here ?

kalitka
Full Member
Full Member
Posts: 14
Joined: Tue Nov 04, 2003 4:13 am

Index Futures End-of-Day Data

Post by kalitka » Sat Feb 21, 2004 6:08 pm

I came up with a promising system for swing trading index futures. I have backtested the system based on indexes downloaded from Yahoo, and it shows quite dramatic results in particular on ^SPX. However, in the upper left corner of Yahoo Finance Page the link to S&P 500 points to something called ^GSPC (what exactly is it ? who computes this index?). My system shows good results there as well but they differ significantly with ^SPX, with returns being inferior. Looks like ^GSPC accounts for gaps as opposed to ^SPX quoted so that Close always equals the next day Open which of course would distort results in real futures trading.
I also have actual daily SP contract data and tested the system on them as well, but my data is not continuous, so for me it's much easier to test on indexes. Is it a grave viloation of "test what u trade trade what u test" tenet? My concerns are intraday futures volatilty, that index data my not reflect. Question: Is historical backtesting on indexes representative for actual index futures trading ? Dividends (reflected in index prices) are also a concern. Anyone can point me to the right direction here ?

Jake Carriker
Site Admin
Site Admin
Posts: 1493
Joined: Fri Sep 12, 2003 10:32 am
Location: Austin, Texas

Post by Jake Carriker » Sat Feb 21, 2004 7:19 pm

Is historical backtesting on indexes representative for actual index futures trading ?
No! Nor is testing continuous contracts and trading actual ones, or testing contracts backadjusted using one method and trading contracts that are backadjusted using another method. Granted, the differences become more apparent when trading longer term, but the basic rule of thumb remains the same: Test what you trade, and trade what you test.

I do not trade intraday, so you might look to others for opinions on that kind of data. I am very satisfied with CSI and their Unfair Advantage software for my end of day data needs. IMHO, data is not a place to pinch pennies. GIGO.

Happy Trading,

Jake Carriker

kalitka
Full Member
Full Member
Posts: 14
Joined: Tue Nov 04, 2003 4:13 am

Post by kalitka » Sat Feb 21, 2004 7:48 pm

Thank you for the answer. I do fully agree with TWYTTWYT.
I should have made myself clearer.
The system under consideration is NOT an intraday system. What I'm trying to do is to use raw DAILY (O,H,L,C) index data to generate signals for actual futures trading - as PART of the system. I will comprehensively test it on continuos futures data (which I do not have at this point) and of course I would trade the same. My reference to futures intraday volatililty was a concern that the stops generated by the system (based on raw index data) may cause excessive whipsaws with futures (as they seem to be more volatile then indexes themselves early in the morning). Still, anyone knows what is GSPC vs. SPX and details on how dividends and gaps are accounted for in either of them ?

Mark_et_Lizard
Roundtable Fellow
Roundtable Fellow
Posts: 62
Joined: Sun Apr 20, 2003 9:16 am

Post by Mark_et_Lizard » Sun Feb 22, 2004 12:43 pm

Never, never, never use index data as the data on which trades are entered in testing. it is completely unrealistic. Trading would be a sweet beautiful dream if actual futures contracts were as neat and orderly as indexes would lead one to believe. All the volatility, all the guts, all the meat of trading is stripped out of indexes.

edward kim
Roundtable Knight
Roundtable Knight
Posts: 344
Joined: Sun Apr 20, 2003 2:42 pm
Location: Silicon Valley / San Jose, CA USA
Contact:

Post by edward kim » Sun Feb 22, 2004 1:28 pm

Mark_et_Lizard wrote:Never, never, never use index data as the data on which trades are entered in testing. it is completely unrealistic. Trading would be a sweet beautiful dream if actual futures contracts were as neat and orderly as indexes would lead one to believe. All the volatility, all the guts, all the meat of trading is stripped out of indexes.
All of the premium or discount that futures often have is stripped-out as well.

I am not sure, but maybe ETFs like SPY and QQQ trade just like the indices. Maybe you can backtest on those instead of the actual index.

kalitka
Full Member
Full Member
Posts: 14
Joined: Tue Nov 04, 2003 4:13 am

Post by kalitka » Sun Feb 22, 2004 4:38 pm

If we are not talking intraday, my understanding is that QQQ and SPY would trade not like indexes but rather almost like futures (they have to -because of arbitrage), the problem with testing on SPY is that its historical quotes have to be adjusted for dividends.

Post Reply