Simple Backtesting

A simple backtester to test trading algorithms and portfolio optimization strategies.

Setup

Docker docker build . && docker run Take note:

• First time this is run, it takes awhile to set up docker environment
• plt.show() wont work. Save it in a file instead (TODO)

Local python -m pip install -r requirements.txt && python -m pip install -e . python loop.py -c credentials.json

Usage

python loop.py -c credentials.json - runs the backtester.

This repo is meant to be as low-level as possible to get greater control of the backtesting environment. Edit the various scripts explained below and import them to loop.py to test your strategies.

Details

This event-driven backtester consists of the following:

• DataHandler (Defines trading universe, reads in necessary data)
• broker Enironment (Details about the exchange that might affect backtesting strategies)
• Portfolio (Executes OrderEvent based on SignalEvent, updates portfolio as necessary)
• Strategy (Looks for SignalEvent to be routed to portfolio)

DataHandler

HistoricCSVDataHandler

Arguments:

• event_queue - An event queue, created as queue.LifoQueue()
• csv_dir - directory where csv files are kept
• symbol_list - List of symbols as stock universe. Ensure that symbols have the same name as the CSV files in csv_dir
• start_date - YYYY-MM-DD. Has to be a trading day else KeyError will be returned.
• end_date (OPTIONAL) - YYYY-MM-DD. Has to be a trading day else KeyError will be returned.

NOTE: Add the above to a github wiki page when more information is available