LLM framework analyzes central bank statements for tone and guidance

Researchers have developed a novel framework using Large Language Models (LLMs) to analyze the sentiment and guidance within Brazilian Monetary Policy Committee (Copom) statements. This system quantifies hawkish and dovish expressions with intensity weights, combining these with document-specific signals to produce a sentiment score ranging from -1 to 1. Additionally, it measures the explicitness and direction of forward guidance, as well as the level and change in uncertainty. The framework was applied to 80 statements from August 2016 to August 2026, revealing that 33.3% of sentences were hawkish and the average document score was +0.107. AI

IMPACT This research offers a new methodological approach for analyzing financial communications, potentially improving sentiment analysis in economic contexts.

RANK_REASON The cluster contains an academic paper detailing a new methodology for analyzing central bank communications using LLMs. [lever_c_demoted from research: ic=1 ai=0.7]

Read on arXiv cs.AI →

• Itaú
• LLM

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →