Abstract
This paper proposes a novel approach of using real time sentiment analysis based on retrieval augmented generation (RAG) for foreign exchange (forex) pricing. It argues that large language models (LLMs) can capture the semantic and emotional aspects of news articles and social media posts that affect the forex market dynamics. Traditional forex trading relies heavily on technical and fundamental analysis. However, the advent of big data and sophisticated text analysis techniques using LLMs have opened up new avenues for analyzing market sentiments in real time that can be extended to Forex. LLMs when used in combination with RAG, an architectural approach to improve efficacy of large language model, can provide real time sentiment analysis. This can be utilized to better manage FX pricing risk in varying frequencies and trading additional settlement currencies from the current once a day model for our card program.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Kurien, Libby A. and Quiroz, Joshua, "A SYSTEM, METHOD, AND COMPUTER PROGRAM PRODUCT FOR DYNAMICALLY ADJUSTING CURRENCIES USING MACHINE-LEARNING MODELS", Technical Disclosure Commons, (September 04, 2025)
https://www.tdcommons.org/dpubs_series/8555