Media and Business Cycle Predictability
Journal of Forecasting · 2026-09 · 45 (6) · 2969-2990
DOI: 10.1002/for.70167 ↗At a glance
Journal article in Journal of Forecasting, published online on 2026-05-07. Written by Salim Baz and 2 co-authors. 41 references.
Abstract (original language)
We construct empirical measures of US business‐cycle activity based on media mentions of the word “recession” in financial newspapers. The MRIs (media recession indicators) are useful predictors of US economic activity, both in‐sample and out‐of‐sample. Moreover, they compare favorably with existing business‐cycle predictors (term premium, purchasing managers' index, consumer sentiment index and real stock market returns). Furthermore, we show that the MRIs are useful predictors of the probability of a US recession 6 months in advance. Our findings also suggest that constructing and using a more sophisticated sentiment‐weighted index does not lead to economically significant improvements in business cycle predictability.
Cite
APA
Baz, S., Cathcart, L., & Michaelides, A. (2026). Media and Business Cycle Predictability. Journal of Forecasting, 45(6), 2969-2990. https://doi.org/10.1002/for.70167
Vancouver
Baz S, Cathcart L, Michaelides A. Media and Business Cycle Predictability. Journal of Forecasting. 2026;45(6):2969-2990. doi:10.1002/for.70167
BibTeX
@article{Baz2026,
author = {Baz, Salim and Cathcart, Lara and Michaelides, Alexander},
title = {Media and Business Cycle Predictability},
journal = {Journal of Forecasting},
year = {2026},
volume = {45},
number = {6},
pages = {2969--2990},
doi = {10.1002/for.70167}
}
EUVIMED