Analyzing NLP for trading insights to assist trading decisions

How ModelOp supports trading decisions at a large US-based broker-dealer

Business Case

The Federal Open Market Committee (FOMC) makes key decisions about interest rates and the growth of the United States money supply. The FOMC meets twice quarterly and releases the minutes from the meetings with a delay.

Though the meeting minutes are released after key decisions have been made (e.g., interest rate increase), the commentary on major macro economic indicators can provide insight into future quarterly decisions by the FOMC.

If one could infer, in real-time, the economic indicators and the accompanying sentiment, you would have an early market advantage in determining future key decisions.

Model Deployed in Business

ModelOp created a real-time, NLP model that identifies major economic topics within FOMC meeting minutes and the associated sentiment on those topics. This model is being used to guide trading decisions for our customer.

Project Specifications:

This ModelOp customer has been built as a cloud-native company, leveraging the Amazon Web Services platform for their core business functions. ModelOp used standard data storage and data access technologies from AWS and extended with custom visualization tools for user interaction.

Model Technique(s)

NLP (spaCy), Logistic Regression, VADER

Model Language

Python

Model Visualizer

Dash

Key Data Sources

FOMC Meeting Minutes

For Businesses, Not Scientists

Thanks to this analytical model, trading decisions can be better capitalize on early knowledge of market forces. This model has also provided an opportunity for our customer to monetize as an analytical product to external customers in financial services.

Explore the Impact of ModelOps

We have the experience and solution to help you harness the power of AI/ML at scale that will boost your core processes. Let’s create a unique solution together.

Contact Brooke for more information