Downloads:
- - Latest version
qsiblive-public-1.0-stable.tar.gz - 17. July 2011 - (README)

- - Earlier versions
qsiblive-public-20110630-0.tar.gz - 30. June 2011 (Order cancellation does not work properly. All earlier versions are also affected by this.)
qsiblive-public-20110608-0.tar.gz - 8. June 2011
qsiblive-public-20110420-0.tar.gz - 20. April 2011
qsiblive-public-20110416-0.tar.gz
qsiblive-public-20110315-1.tar.gz



What is it?   'qsiblive' is a collection of R functions that will enable you to use a trading strategy developed in the backtesting evironment 'quantstrat' for live trading with interactive brokers. This code is in constant development , it has not been optimized nor has it been cleaned of unnecessary comments and should NOT be used with a production (=CASH) IB account. I recommend that you use it for test-trading with an IB PAPER account only. It is built on top of two great R packages,'quantstrat' and 'IBrokers'. For details, see the README.

What can it do?   It can show you how to deploy a tested and hopefully robust and profitable trading strategy to the market, with R alone. This version is limited to End-Of-Day trading based on daily Close prices. For now it only supports trading a single instrument, not a portfolio of instruments. While in theory you can launch as many instances of it as you like and therefore effectively trade a portfolio of single instruments, using either the same strategy or a different one on each instrument, I have not tested this. 'qsiblive' is not supposed to be a fire-and-forget solution. It should definitely not be used for unsupervised trading. But I think it can give you an idea how to create your own solution. One that you feel comfortable with and trust.

Why is this useful?   Well, there are many nice pieces of software that connect to the IB Trader Workstation,some of which can be used for automated trading. Some are even freeware. None or very few of them however offer the power, simplicity and community support of the 'R' project. Besides, R has really an amazing repository of free and very adavanced statistical packages, something you will find very useful if you trade using quantitative methods. All I am trying to do here is bring R a bit closer to the market, without any middleware.

Quote: "It's not pretty, but it works"


Soren Wilkening (Dipl.Math.)
me (arroba) censix (punto) com
LinkedIn