Is Bitcoin A Hedge Against Central Banking? Evidence from AI-Driven Monetary Policy Expectations

Maxime L. D. Nicolas, Fran\c{c}ois Sicard, Marion Laboure, Zixin Sun and Anah\'i Rodr\'iguez-Mart\'inez

Papers from arXiv.org

Abstract: This study investigates the transmission of monetary policy narratives to Bitcoin prices, distinguishing policy expectations from realized policy implementation. We introduce a weekly Monetary Policy Expectations (MPE) index derived from the Large Language Model (LLM)-based classification of 118,000+ market messages, providing a granular measure of hawkish and dovish monetary policy discourse. We demonstrate that changes in the MPE index provide evidence of significant linear predictive information for Bitcoin returns at short-to-medium horizons, with significant Granger causality at multiple lags. A Long Short-Term Memory (LSTM) framework combined with SHapley Additive exPlanations (SHAP) further identifies nonlinear and regime-dependent relationships between monetary-policy expectations and Bitcoin returns, indicating that Bitcoin functions as a sensitive barometer of central bank signaling. In particular, hawkish monetary-policy narratives are associated with negative price responses that are not accounted for by contemporaneous Federal Funds Rate adjustments. These findings highlight Bitcoin's structural sensitivity to global monetary discourse, establishing LLM-derived monetary-policy sentiment as a high-frequency measure of central-bank communication and as an informative leading macroeconomic indicator for the digital asset landscape.

Date: 2026-04, Revised 2026-08
New Economics Papers: this item is included in nep-big, nep-cba, nep-fdg, nep-fmk, nep-mon and nep-pay
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
https://arxiv.org/pdf/2604.08825 Latest version (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:arx:papers:2604.08825

Access Statistics for this paper

More papers in Papers from arXiv.org
Bibliographic data for series maintained by arXiv administrators ().