This repo contains utilities for developing and backtesting algorithmic trading strategies on Manifold Markets. It is designed to be used with my fork of Manifoldpy. I wouldn't suggest using the original Manifoldpy, as I've made changes to make it easier to work with Manifoldpy locally. You're welcome to try, though.
The main contributions of this repo are strategy.py
and time_utils.py
. The former contains a base class for developing and backtesting strategies, and the latter contains utilities for working with time series data. Once you've looked over those, feel free to try out your own strategies in strategy_anal.ipynb
(naming deliberate.)
If you want to get a sense of how the Strategy
class works in practice, check out some of the example strategies in the example_strategies
folder!
https://github.com/jagilley/manifold-strategies && cd manifold-strategies
git clone https://jagilley/manifoldpy
pip install tqdm pandas openai