Best Historical ES Tick Data?

T

Truckload

I would like to find a source of continuous contract tick data for the ES that I can then use with Ninja Trader. Any ideas.

I also use Tradestation and it provides about 6 to 9 months of tick data on the continuous ES contract. However, I am having trouble getting that data into Ninja. Any ideas?

Thanks.
 
You can buy tick data directly from the CME. It won't be continuous contract, but there are well known ways to create continuous contract pricing.

Best regards,

Louis
 
Hi, can you please tell me where on cme's website you can buy the historical tick data (just normal time & sales) for es mini.

It has the historical data for free going back a max of 5 days only. I am looking for historical es tick data going back 3-5 years, probably a few gb worth.

How can i buy this from cme? I guess they have to provide the front month contract for each of the active trading period of that contract, so i can combine them into 1 continous feed for backtesting.

Thank you for your help!
 
Quote from newguy05:

Hi, can you please tell me where on cme's website you can buy the historical tick data (just normal time & sales) for es mini.

It has the historical data for free going back a max of 5 days only. I am looking for historical es tick data going back 3-5 years, probably a few gb worth.

How can i buy this from cme? I guess they have to provide the front month contract for each of the active trading period of that contract, so i can combine them into 1 continous feed for backtesting.

Thank you for your help!
More...

http://www.cme.com/trading/dta/hist/hftp/histftp.html
 
Thanks i found it, tick data for es from 1997 till now is $3500, a bit more expensive than i expected, but the main problem is the file size, it's not a few GBs but 147GB! I guess they are giving you all the tick data for all the contracts regardless of front month or not.

I was going to write a java program to parse this into the database for backtesting, but parsing 147gb worth of data is too much. Hmm....
 
Quote from lrm:

You can buy tick data directly from the CME. It won't be continuous contract, but there are well known ways to create continuous contract pricing.

Best regards,

Louis
More...

Could you please provide some links or documentation on how to create the continous contact pricing.

Appreciated.

NN
 
I was looking at tickdata.com recently and IIRC the whole tick history for SNP was cheaper than that.

And I've never seen any Java programs that create continuous contracts. Presumably you mean back-adjusted as well. If something did exist, it would have to have the right input and output formats for your software. Sounds like a big ask for open-source shareware, although not unimaginable.
 
Quote from ahardy66:

I was looking at tickdata.com recently and IIRC the whole tick history for SNP was cheaper than that.

And I've never seen any Java programs that create continuous contracts. Presumably you mean back-adjusted as well. If something did exist, it would have to have the right input and output formats for your software. Sounds like a big ask for open-source shareware, although not unimaginable.
More...

Yes i found plenty of websites selling the es tick data for only a few hundred bucks, the problem is i really dont know how accurate they are. Your back testing is only as good as your data, i know cme's data is 100%...maybe i am just being paranoid.

The java program is somethng i will write from scratch myself, it just adjusts the contracts to take take 1 contract (the front month) for each month, and load the data from whatever format the vendor file is in into the database.

Once it's in the database everything becomes trivial, as db are designed for data querying, put a clustered index on the filter columns and you can code any back testing you want.

I wasnt able to find any backtesting software that can code my strategy easily, tradestation came pretty close but it's very messy. I am trying to backtest a point&figure strategy, and that makes things complicated as most backtesting softwares are designed to work with time series only.

The problem is loading the 147 GB of data into the database using java. It will take forever. Maybe i will just deploy it on some remote server and let it run...
 
147 Gb is huge - but I guess it's just ASCII text.

Maybe you can erase the data you've read as you go along, so that you don't have to do it all in one hit. Not that I have any experience in deleting from the front of files without rewriting the whole file!
 
You must log in or register to reply here.
Back
Top