April 25, 2023
A recent study on Forex trading shows combining RavenPack news sentiment and fundamental input variables in a Machine Learning model improves predicting USDBRL's 21-day return.
The study by Alex Giulietti de Barros, Felipe Honório and Emerson Fernandes Loureiro showed that sentiment data has predictive power when applied to forex trading over a 21 days horizon. The news sentiment was pre-processed and combined with fundamental data, and a basket of emerging market currencies.
In particular, they found that:
The sentiment-enhanced model outperformed the benchmark in every year between 2016 and 2020, with a Sharpe Ratio of 2.27 vs. 0.05 for the benchmark.
Results showed that RavenPack news sentiment was able to enhance the strategy performance, both relative to a buy-and-hold strategy, as well as to a pure fundamental model.
Please use your business email. If you don't have one, please email us at info@ravenpack.com.
By submitting this form, you agree to RavenPack's terms of service and privacy policy.
Your request has been recorded and a team member will be in touch soon.