The National Bureau of Economic Research (NBER) recently published a study by Isha Agarwal from the Sauder School of Business at the University of British Columbia, and Wentong Chen and Eswar S. Prasad from Cornell University, examining the role of media-driven narratives in shaping cross-border institutional investment flows, with a focus on China. By analyzing over 1.4 million news articles published in 15 countries from 2007 to 2022, the researchers utilized natural language processing (NLP) techniques to construct sentiment and risk indices that captured the tone and perceived investment risks portrayed in local media. These indices revealed how domestic media narratives influence institutional portfolio flows into China, even when controlling for macroeconomic and financial fundamentals, and highlighted substantial variation in sentiment and risk perceptions across countries.
The study's findings underscored the significant role of media-driven narratives in driving cross-border capital flows. By correlating changes in sentiment and risk indices with fund-level data, the researchers demonstrated that media narratives strongly influence investment decisions. For instance, a one-standard deviation increase in positive sentiment about China corresponded to a 2.3% rise in quarterly portfolio inflows, while a similar rise in perceived risk was linked to a 2.1% decline in investments. These effects were particularly pronounced in times of heightened uncertainty when interpreting traditional market data became more challenging. Investors with greater familiarity or private knowledge of China, such as those with significant prior exposure or geographical proximity, were less influenced by media narratives. However, newer or less familiar investors relied more heavily on these narratives, especially during volatile periods. This reliance often reflected the higher costs and challenges of accessing reliable and timely data about China's opaque regulatory and economic environment.
The research also revealed a nuanced relationship between narrative categories and investment flows. Political and environmental narratives, often considered secondary to economic ones, were shown to have as much, if not more, influence on investment decisions. This phenomenon is partly attributed to the lack of standardized data and the difficulty of interpreting information on non-economic factors. For example, political narratives related to policy uncertainty or environmental coverage on China's green initiatives provided critical context for investors navigating complex decisions. The study also showed that investors react more strongly to negative media coverage than to positive stories, amplifying the impact of the media's inherent negativity bias. This asymmetry suggests that investors are more cautious and likely to pull back investments in response to adverse narratives, reflecting a broader tendency to prioritize risk aversion over opportunity-seeking.
Another key finding was the variation in media sentiment and risk perceptions across countries. Local media coverage often diverged significantly in tone, even when reporting on the same event. For instance, U.S. media outlets framed China's 2016 stock market crash with terms like "panic," whereas Singaporean media adopted a more neutral perspective. Such disparities were linked to domestic biases, geographical proximity, and editorial decisions. The research highlighted that media narratives are not merely reflections of current or anticipated economic conditions in China but also represent broader interpretations that can diverge from underlying market fundamentals. This divergence underscores the role of local media as a powerful intermediary shaping investor sentiment and risk perceptions.
To better understand the mechanisms driving these results, the study explored how different types of investors responded to media narratives based on their information access and familiarity with China. Investors with early exposure to Chinese markets or larger holdings in Chinese assets were less sensitive to media sentiment, indicating that private or direct knowledge reduced their reliance on external narratives. Conversely, during periods of high market volatility, all investors showed increased sensitivity to media narratives, illustrating the heightened role of narratives in uncertain times. This suggests that media narratives play a dual role: serving as both a lens through which less informed investors evaluate opportunities and risks and a fallback source of information during periods of heightened uncertainty.
The research provides actionable insights for policymakers and investors. Policymakers in emerging markets like China could reduce investor reliance on media narratives by enhancing data transparency and accessibility. Publishing economic and regulatory information more frequently and in multiple languages could help offset the influence of potentially biased media coverage. For investors, recognizing the impact of local media narratives can guide better decision-making and mitigate risks associated with overreactions to negative coverage. Additionally, this study emphasizes the need for global investment strategies to consider the broader socio-political and environmental narratives that influence market dynamics beyond traditional economic indicators.
By focusing on the interplay between media narratives and global financial flows, the study adds to the growing body of literature on narrative economics and cross-border investments. It extends previous research by demonstrating how media narratives influence open economies, particularly in the context of information frictions and emerging markets. Furthermore, the findings reveal the importance of considering diverse narrative categories economic, political, and environmental in investment decision-making. The results highlight the broader implications of narrative-driven biases in shaping investment behaviors, offering a deeper understanding of how media acts as both an information source and a sentiment driver in global capital markets.
{{#Source}}{{Source}}{{/Source}}{{#IsBlog}}
{{Disclaimer}}
{{/Disclaimer}}