Enhancing Regime Shift Detection Using Unstructured Data: A Study on the Treasury Market

2026-05-17

arXiv.org

AI Summary

This study addresses the challenges in identifying financial market regime shifts—namely, high data noise, strong multicollinearity, and underutilized unstructured textual information—by proposing a text-augmented dual-channel detection framework. The approach uniquely integrates FOMC meeting minutes, from which potential regime-switching points are extracted using a large language model, with multivariate time series analyzed via vector autoregression (VAR) likelihood ratio tests. By cross-validating signals from these heterogeneous sources, the framework achieves a detector-agnostic architecture capable of flexibly incorporating any data-driven method. Evaluated on U.S. Treasury market data from 2010 to 2024, the proposed method attains an F1 score of 0.82 and an F2 score of 0.86, substantially outperforming purely data-driven baselines while enabling same-day detection latency.

Abstract

Regime shifts in financial markets reorganise the joint dynamics of asset prices and macro variables, breaking any single-regime calibration. They are nonetheless difficult to detect reliably because the data signal is noisy and heavily multicollinear, while the contemporaneous text that announces them is unstructured. Standard regime shift detection methods rely solely on structured time-series data and ignore policy communications, even though these texts often signal shifts before they materialise in observed prices. We propose a text-enhanced regime shift detection pipeline that combines large language model (LLM) reasoning over central-bank communications with statistical validation on multivariate financial time series. The framework is detector-agnostic: text-proposed candidates are validated using a bootstrap likelihood-ratio test on a vector autoregression (VAR), while data-driven candidates from arbitrary regime detectors are ratified through a lenient LLM text check. We evaluate the framework on 2010-2024 FOMC minutes paired with a 14-variable U.S. Treasury and macroeconomic panel, using four interchangeable data-driven detectors. The proposed pipeline achieves F1 = 0.82 against a verified anchor list of monetary-policy regime shifts, with same-day modal detection latency and consistently stronger performance than pure data-driven baselines. The results demonstrate that combining unstructured policy text with statistical structural-break detection improves the robustness and interpretability of regime shift identification in financial markets.

Problem

• regime shift detection
• unstructured data
• financial markets
• Treasury market
• monetary policy

Innovation

• regime shift detection
• large language model
• vector autoregression
• unstructured text
• detector-agnostic