Reproduce “Mind the Shift: Decoding Monetary Policy Stance from FOMC Statements with Large Language Models” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Yixuan Tang, Yi Yang
DCS uses consecutive FOMC statements to learn monetary-policy stance without stance labels, combining frozen language-model representations with a consistency penalty linking each statement’s score to its change from the prior meeting. On an expert-annotated sentence benchmark, Qwen3-4B DCS achieves 71.08% accuracy, versus 61.45% for a supervised linear probe; inflation and Treasury-yield results are associations, not causal effects.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.