backtest-pro is a framework that provides a way to test complex strategies in an environment that is designed to look as much as possible to the real world. This way, the historical results are more likely to reflect the real world results. It is an event-driven backtesting framework that is designed to be as flexible as possible and as complete as possible. It supports end-to-end quant pipeline from data fetching to production release.
Also, it has the broader goal of becoming the most complete backtesting framework available for python finely tuned for professional applications.

backtest-pro is still in development and is not ready for production use. There may be bugs that could make the results of the backtest invalid. Always double-check the results with your own code. If you find a bug, please open an issue on the github page. The api might also change without notice.

• Main Features
• Installation
• Installation from source
• License
• Documentation
• Contributing
• Beware
• RoadMap

Here are just a few of the features that backtest-pro offers:

• DataPiplines
  • An easy to use data pipeline api that makes building a data pipeline a breeze.
  • A pipeline built with the api is easy to maintain and easy to understand.
  • Support caching for more efficient pipelines.
• Backtest
  • Backtest with a single or with multiple assets simultaneously
  • Feed a moving window to the strategy
  • Multiple time resolution simultaneously
  • Take into account:
    • Trading fees
    • Margin rates
    • Margin calls
    • Stock splits
    • Dividends
  • Records a lot of metrics for easier analysis and debugging.
• Release
  • Use the same code as used in the backtest in production.
  • Only a few lines of codes are necessary to build a production pipeline that can run on a server or locally.
  • Automatic reporting of the results using report builders. (Html, pdf)
  • Easy integration with other services such as api for algorithmic trading.

To install backtest-pro, you can use pip:

pip install backtest-pro

There a few dependencies that are not installed by default. They are:

• TA-Lib: A technical analysis library that is used to calculate technical indicators.
• Plotly: A plotting library that is used to render charts in the production of reports.
• WeasyPrint: A library that is used to convert html to pdf. It is used to render the reports in pdf format.
• kaleido: An optional library of Plotly that is used to render the charts in the reports.
• schedule: A library that is used to schedule the run of the strategy in production.
• python-crontab: A library that is used to schedule the run of the strategy in production.

To install the dependencies, you can use the following command:

pip install backtest-pro[optional]

To install backtest-pro from source, you can clone the repository and install it using pip:

git clone https://github.com/anthol42/backtestPro.git

Move to the cloned repository:

cd backtestPro

Then, you can install with the following command:

pip install .

GNU General Public License v3.0

Click here

All contributions are welcome. It can be a bug report, a bug fix, a new feature, a documentation improvement, etc.
If you want to contribute, please read the CONTRIBUTING.md file.

Do not forget to add tests for your code and to run all the tests before submitting a pull request.

As contributors and maintainers to this project, you are expected to abide by pandas' code of conduct. More information can be found at: Contributor Code of Conduct

There are a few things to keep in mind when backtesting strategy to avoid unrepresentative results:

• Survivorship bias: Make sure to include all the assets that were available at the time of the backtest, including delisted assets.
• Look-ahead bias: Make sure that your data pipeline does not include data that would not have been available at the time of the backtest. This could be, for example, normalizing the data with the mean of the whole series or dataset. It could also be to fetch fundamental data, but not using an appropriate padding, which would cause the fundamental data to be shifted from the price data. If the data is fetch and preprared correctly, the backtestPro framework will make sure the strategy won't use future data.
• Split Bias: Make sure to include the stock splits in the data. The backtestPro framework will take care of the stock splits.
• Dividend Bias: Make sure to include the dividends in the data. The backtestPro framework will take care of the dividends.
• Trading fees: Make sure to include the trading fees in the backtest. The backtestPro framework will take care of the trading fees if configured correctly (passing the trading fees to the backtest object during initialization).

Reference: https://seekingalpha.com/performance/quant (Accordion: What are common issues with back testing and performance results?)

• Alpha Release
  • Code implementation
  • Documentation (In progress: using pdoc3, search bar)
  • Notebooks tutorials (Missing few tutorials: Serve module and Indicator module)
  • Make documentation public
• Beta Release
• Stable Release