Documentation

Onboarding

Welcome to the team! Please first read the following document to gain a good understanding about our trading system.

Basic knowledge

Basic finance terminologies, Investopedia is your bestfriend. Read definitions on Investopedia and do a brief research before asking.
Watch the following video from MIT opencourseware
Basic knowledge in calculus, statistics and probability. In the future, we may work on stochastic calculus and differential equations. Experience in math modeling is recommended.
Documents will be given to people working on specific areas. Contact Hoang and Long for more details.

What we do

As we are yet to have an official name, we will refer to this system as "Tradingsystem". Tradingsystem is an algorithmic trading system, but we don't limit ourselves to quant as we also do technical analysis and fundamentals analysis. At the moment, we are trading forex and we may expand to stocks and derivatives in the future.

Algorithmic trading is an extremely rewarding academic-wise. We have several projects in software engineering, economics, finance mathematics and machine learning for everyone to work on. Here, we have a list of open tasks.

Infrastructure Description:

Our system mostly uses Python 2. It is advised for newcomers to have good knowledge in Python and feel comfortable in using GitHub.

Threads:

streaming.py: output - real time EURUSD price from Oanda as TICK events (check event.py)
strategy.py: analyze TICK events to determine when to enter. Output is SIGNAL events.

Strategies:

See strategy folder. There are three main strategies:

Mean-reversion
Momentum
Moving averages

Risk management:

portfolio.py: receive SIGNAL events - conduct risk management. Output - ORDER event. (under dev)

Simulation - Backtesting:

Expectation: modify strategy.py --> strategy_session.py. User could choose to backtest or trade live. We would import BacktestBase.py, which is under development. The expected output is statistical indicators of strategy's performance.

hist.py: download historical data for analysis and building backtests.

Evaluation:

statisticaltest.py: statistical analysis to evaluate strategy's performance.

Dependencies

Python 2.7. PyPy JIT Compiler 5.8.0.

All required packages are listed in requirements.txt.

Installation

These software/packages must be installed in sequence. We haven't prepare documentation for Windows users.

Oracle Java Development Kit 8

Ubuntu

Visit here for more details.

macOS

Install HomeBrew.
$ brew cask install java
Open file ~/.bash_profile and add these two lines:

export JAVA_HOME=/Library/Java/JavaVirtualMachines/1.x.x_xxx-bxx
export PATH=$JAVA_HOME/bin:$PATH

Apache Cassandra

Download Cassandra here. Unzip the downloaded file.
Open ~/.bash_profile (on macOS) or ~/.bashrc (on Ubuntu).
Add these two lines:

alias cassandra='p/a/t/h/apache-cassandra-x.xx/bin/cassandra'
alias cqlsh='p/a/t/h/apache-cassandra-x.xx/bin/cqlsh'

Activate the alias by restarting Terminal or source ~/.bashrc (Ubuntu) or source ~/.bash_profile (macOS).

PyPy

Download PyPy for Python 2.7 here.

VirtualEnv

$ [sudo] pip install virtualenv
Inside the folder Tradingsystem, run $ virtualenv -p p/a/t/h/pypy2-v.5.71-osname/bin/pypy venv.

Contribution

Fork this repository to your own GitHub account.
Download the project to your machine: $ git clone https://github.com/username/Tradingsystem
Follow the installation guidance.
Get it run:

$ source venv/bin/activate
(Tradingsystem) $ pip install -r requirements.txt
$ cassandra
$ python __main__.py

Other

Contact long.le@minerva.kgi.edu and hoang.nguyen@minerva.kgi.edu for more details. Task distribution will be according to project list google sheets file.

LongntLe/Tradingsystem