We study market integration through the lens of how sentiment extracted from news affects aggregate market returns. More specifically, we investigate whether sentiment on a given country contains predictive information about local and international market returns with trading strategies.
To do so, we employ a novel dataset consisting of 520 million articles from 2015 to 2023 across 14 developed countries. Our findings reveal a strong local effect, especially for the U.S., where news from a country about itself significantly predicts its market returns. The U.S. also demonstrates substantial influence as a source and target of news across nine of the fourteen markets analysed, suggesting that developed countries are integrated with U.S. markets. Additionally, we highlight significant market integration within European countries, with German news exerting predictive power over other E.U. markets. This is in contrast to Asian countries, suggesting regional differences in market dynamics.
