Parsing the pulse: decomposing macroeconomic sentiment with LLMs
Summary
Focus
The rise of large language models (LLMs) allows researchers to capture context, nuance and subtle meaning from text. This was not possible in the older generation of textual analysis. Potential macroeconomic applications are vast but remain relatively unexplored.
Contribution
We apply LLMs to a large corpus of US news articles to construct macroeconomic sentiment indices. We decompose these into demand and supply components, as well as more granular drivers. This approach delivers near real-time measures of growth and inflation, accompanied by a narrative of the underlying forces.
Findings
The news-based sentiment indices closely track conventional hard-data benchmarks. The underlying components are intuitive and consistent with prior studies. Adding sentiment indices to forecasting models improves out-of-sample performance. The results highlight the value of LLMs as a complement to traditional model-based approaches to structural analysis.
Abstract
Macroeconomic indicators provide quantitative signals that must be pieced together and interpreted by economists. We propose a reversed approach of parsing press narratives directly using Large Language Models (LLM) to recover growth and inflation sentiment indices. A key advantage of this LLM-based approach is the ability to decompose aggregate sentiment into its drivers, readily enabling an interpretation of macroeconomic dynamics. Our sentiment indices track hard-data counterparts closely, providing an accurate, near real-time picture of the macroeconomy. Their components–demand, supply, and deeper structural forces–are intuitive and consistent with prior model-based studies. Incorporating sentiment indices improves the forecasting performance of simple statistical models, pointing to information unspanned by traditional data.
JEL classification: E30, E44, E60, C55, C82
Keywords: macroeconomic sentiment, growth, inflation, monetary policy, fiscal policy, LLMs, machine learning
