This project provides a CI/CD framework to design, test, and implement algo-trading robots using Interactive Brokers and Python. The project wraps the TWS_API into a new API that manages signal subscriptions and orders in a layered approach.
The major benefit of this approach is that the TWS_API adapters can be replaced with back-testing adapters. Interactive Brokers, unfortunately, does not provide a back-testing platform and only provides a paper account that operates like a live account for testing – meaning that back-testing cannot be done out-right with TWS_API. This project provides these back-testing adapters.
This means that we can now train and test algo-trading robots on historical data through this framework, and we merely have to change the adapter once we are ready to test the robot in the paper account, and subsequently ramp it up in the live account.