Title : Econometric Forecasting Using Ubiquitous News Text: Text-enhanced Factor Model

Author : Beomseok Seo(BOK)

The use of news text as a novel source for econometric forecasting is gaining increasing attention. This paper revisited the way of incorporating narrative information into econometric forecasting by effectively quantifying sector-specific textual information without requiring training data. We exploit Theme Frequency Indices(TFI) utilizing domain-specific subject-predicate patterns to gauge public perception of the economy. TFIs of 15 sectors, including production, inflation, employment, capital investment, stock and house prices, and others, were examined and integrated into Text-enhanced Factor Model(TFM) using latent factor structures. Empirical analysis, based on over 18 million news articles in Korea, reveals that TFM improves the accuracy of near-term GDP forecasts, demonstrating simple text-mining techniques along with domain knowledge are capable of leveraging qualitative information from news without costly training. The proposed method is applicable to a wide range of subjects for measuring narrative information of the economy, offering a rapid and cost-effective approach.