FinSentGPT: A universal financial sentiment engine?
Mahdavi Ardekani, A., Bertz, J., Bryce, C. ORCID: 0000-0002-9856-7851 , Dowling, M. & Chen, S. (2024). FinSentGPT: A universal financial sentiment engine?. International Review of Financial Analysis,
Abstract
We present FinSentGPT, a financial sentiment prediction model based on a fine-tuned version of the artificial intelligence language model, ChatGPT. To assess the model’s effectiveness, we analyze a sample of US media news and a multi-language dataset of European Central Bank Monetary Policy Decisions. Our findings demonstrate that FinSentGPT’s sentiment classification ability aligns well with a prominent English-language finance sentiment model, surpasses an established alternative machine learning model, and is capable of predicting sentiment across various languages. Consequently, we offer preliminary evidence that advanced large-language AI models can facilitate flexible and contextual financial sentiment determination, transcending language barriers.
Publication Type: | Article |
---|---|
Additional Information: | © 2024. This manuscript version is made available under the CC-BY-NC-ND 4.0 license https://creativecommons.org/licenses/by-nc-nd/4.0/(opens in new tab/window) |
Publisher Keywords: | ChatGPT; large language models; financial sentiment; monetary policy; fine- tuning |
Subjects: | H Social Sciences > HG Finance Q Science > QA Mathematics > QA75 Electronic computers. Computer science |
Departments: | Bayes Business School Bayes Business School > Actuarial Science & Insurance |
SWORD Depositor: |
This document is not freely accessible due to copyright restrictions.
Available under License Creative Commons Attribution Non-commercial No Derivatives.
To request a copy, please use the button below.
Request a copyExport
Downloads
Downloads per month over past year