We’re proud to announce, thanks to the support of the community, the LEAN Algorithmic Trading Engine is now 100% open source. You have the freedom to connect any data source, execute through any brokerage and design any algorithm 100% locally.

Moment of our Open Sourcing, Jan 12th 2015

It’s an exciting new frontier for algorithmic trading; through open source QuantConnect is breaking open the traditionally secretive world of algorithmic trading togive you the same powerful tools as major hedge-funds.

Lean is “plug and play“.Running your first backtest takes about 23seconds.

1. Star/ForkandDownload the QuantConnect/LEAN Repo* fromGitHub
2. Open Lean Project in Visual Studio (let Nuget download all dependencies)
3. Press F5 to Run Project

Presto-you’ve run your first backtest! Hereis a step by step guide for building your first algorithm. You can alsodesign custom indicators, importdata for international stock markets and connect with any brokerage. We even ship some data with the repo so you can get started instantly.

Clone LEAN Today to Start Your Journey

Clone LEAN Today to Start Your Journey

We’re incredibly grateful to the QuantConnect pioneers for making this possible.With your support we can build the best algorithmic trading platform in the world. Sustainable, independent and community driven.

More Raw Power

To be profitable you need to iterate quickly. Last weekwe upgraded our backtest processing servers:you can now run a 10 year, event driven backtest in 33 seconds. Your algorithms are running on beautiful i7x3930’s with 6 cores/12 threads/64GB ram.Weare the world’s first cloud-desktop hybrid algorithmic trading platform aiming to give you the best of both worlds;ease of local development and horse power of the cloud.

Dynamic Indicator System

Thanks to some long hours by Michael H we launched an elegant, powerfulanddynamicnew indicatorlibrary.It lets you implement designs quicklyand avoids reinventing the wheel.Creatinganindicator is onlya single line of code! Get started with the sample algorithm.

var rsi = RSI("SPY", 14);
var bb = BB(_symbol, 20, 1, MovingAverageType.Simple, Resolution.Daily);
if (rsi > 80) {
 SetHoldings("SPY", 1);
} else if (rsi < 20) {
 SetHoldings("SPY", -1);
}
Plot("BB", bb.UpperBand, bb.MiddleBand, bb.LowerBand);
Clone the sample algorithm which implements 15+ indicators.

Bollinger Bands Implementation - Clone the sample algorithm which implements 15+ indicators.

 

Comments


  1. George says:

    Ironically, in order to use this 100% open source code, you need Micro$oft tools.

  2. Jared Broad says:

    LEAN was actually built for Linux/Mono – but thankfully Microsoft have also open sourced Visual Studio! 🙂 http://www.visualstudio.com/products/visual-studio-community-vs

  3. Marc says:

    Is it possible to run on Mac/Linux distros?

    1. James Crocker says:

      Yes. I run on Ubuntu 14.04 64bit with Mono .NET and Mono Developer for IDE. There are a couple of NuGET packages to load; but nothing extraordinary to get running. If needed I can post something in the Community Discussion Forums.

  4. Jared Broad says:

    @Marc – I believe so but have never done it myself (http://www.mono-project.com/download)

  5. Tom says:

    In the manual for lean trading engine there is a picture about the higher architectural view. When I click it it leads me to a github page that is not loadable.. The image is too small so can’t see it properly. Any chance you can fix that?

    Another question I had: does the lean engine have the ability to do portfolio backtests? For example, 1 strategy with many symbols or multiple strategy on 1 symbol or multi-strategy multi symbol?

    thanks and great contribution to open source.

    1. Jared Broad says:

      Thanks Tom appreciate the support! I’ve updated the documentation. The raw link is here in github.

      We can definitely do portfolio backtesting – 1 strategy with many symbols is done. If you needed multiple strategies running in parallel you can launch multiple live trading servers which all act independently – however your brokerage can only support one login. You can still do multiple strategies in one algorithm you’d just need to manage which one takes priority internally.

  6. Great news Jared !

    @George M$ has been opening up for the last 5-6 years and you can get VS2013 Pro for free by downloading the community version (no limitations aside from licencing). Also C#, MVC, .NET, Entity Framework and many other things are all open source to which you can contribute on Github if you wish.

  7. Markus says:

    This is really great!
    Thanks a lot.

  8. Umesh Chand Sharma says:

    Sorry being Dumb! Do I need to download Lean along with Visual Studio Community?

    1. Jared Broad says:

      If you already have Visual Studio you won’t need to download the community edition. The only other external requirement is NuGet which comes with visual studio these days I believe.

  9. Will the rest of the QuantConnect platform become open-source?

    1. Jared Broad says:

      Everything needed for local algorithmic trading has been open sourced; including connecting to your own brokerages and data sources. We made the decision that code which runs our website & cloud (JS-IDE, compiler, auto-complete) isn’t useful for most people’s algorithmic trading – and isn’t part of the Lean engine.

      1. Markus says:

        The Visual Studio UI plugins would still be very useful and they would not be a part of the website.

  10. aadon says:

    Can you use F-Sharp(F#) with this?

    1. Jared Broad says:

      I believe so yes; I’m just not sure how to do the method calls for a F# algorithm.

Sign In to comment in this post.