Measuring economic outlook in the news

TL;DR

This paper introduces a resource-efficient method combining document embeddings and synthetic data from large language models to measure economic outlook from news, improving forecast accuracy and sentiment detection.

Contribution

It presents a novel, interpretable approach that enhances economic sentiment analysis using minimal resources and proprietary news data, outperforming traditional methods.

Findings

Significantly improves GDP growth forecast accuracy

Detects sentiment shifts weeks before official releases

Outperforms survey-based and dictionary methods

Abstract

We develop a resource-efficient methodology for measuring economic outlook in news text that combines document embeddings with synthetic training data generated by large language models. Applied to 27 million news articles, the resulting indicator significantly improves GDP growth forecast accuracy and captures sentiment shifts weeks before official releases, proving particularly valuable during crises. The indicator outperforms both survey-based benchmarks and traditional dictionary methods and is interpretable, allowing identification of specific drivers of economic sentiment. Our approach addresses key institutional constraints: it performs sentiment classification locally, enabling analyses of proprietary news content without transmission to external services while requiring minimal computational resources compared to direct large language model classification.