Traders' Roundtable

ok - I have to admit this is covering ground from some earlier posts/topics. However I'm still a little confused so if someone can make the responses to the below 3 questions as spoon-fed as possible that would be great!

a) As a non-programming techno-phobe, what is the best way for me to get a fully comprehensive historical library of whichever prices I want to use in Trading Blox without having to manipulate the data in any way? MetaStock's Reuters DataLink have an offer where for just over $100, a single monthly subscription feed of all their data going back years and years is available. Is their data a good fit for TB?

b) What's the idiot's guide explanation to the futures roll issue and again how does this impact choice of backtest data? Is there a clear winner if this variable is added into the mix?

c) If a) +b) above leads to CSI as the best option and I need to wince and pay up $360 in licence fees and $190 in monthly fees (for NA Futures + World Futures + Single-Stock-Futures, US Stocks and Indices + Non-US Stocks and Indices), is it correct that it is Rent-a-Data so the monthly charge has to be paid every month even to access the historic data? So I can't pay the licence fee + 1 monthly fee and then test for say 3 months until ready to choose which markets I want to trade live at which point I take out my targeted subscriptions?

I really want to get going on system building but this is proving to be a roadblock...

Hi,

You need to think of it as an investment. If shelling out a few hundred dollars is going to enable to get you thousands in return, then whats the problem ? Don't take it personally - its just business. All companies (from the smallest to the largest) have to invest something up-front to get started.

I don't see the massive issue here. If you've got enough to trade futures, you'll make the years subscription back easily in a single profitable trade.

As for rolls - just roll on open interest (first or second trigger) for all your testing. Then when you're satisfied with performance etc, spend a little time collating the First Notice Day for each of the contracts you're interested in & play with the roll date (either fixed or OI based) to make sure you roll before FND but on good volume/OI. I don't think the exact roll timing makes a huge difference to system performance of a Trend Following System.

Easy route ?? Just go for CSI. Integrates easily with TB, lets you play with rolls very simply with no coding. Then you can concentrate on what's important (i.e. your system)

Dave

I think Dave makes some very good points. There is absolutely no substitute for accurate and reliable market data as a fundamental building block of your system. It is not glamorous but absolutely necessary. You can lose more than your annual subscription in a flash on some bad trading decisions with data integrity issues.

I use CSI (based on guidance from this forum) but I would recommend that you take a look on this forum for some information about how others find CSI and how they have worked around some of the integrity issues. I would in particular start with a post that includes a link to some TB code that checks the integrity of the market data from CSI (or any vendor for that matter) kindly supplied by Sluggo.

viewtopic.php?t=4003&highlight=

... in case you missed it in your search of this forum.

Good luck