Foreign gatekeepers for local markets

Technical Summary: Foreign Gatekeepers for Local Markets

Problem Statement
The paper investigates how information available to global investors influences local market dynamics in emerging markets (EMs), specifically focusing on India. The central problem addresses the deterioration of local information channels due to regulation, censorship, or systematic bias, which increases information frictions and undermines cross-border capital allocation efficiency. The author asks whether foreign media can substitute for distorted local information and what the consequences are for market dynamics. While foreign media might lack granular local details, the paper posits that they may act as "gatekeepers," shaping expectations and market-wide dynamics by filtering aggregate and idiosyncratic information. The study is motivated by recent regulatory pressures on Indian media (e.g., the 2020 Amnesty International exit, 2023 BBC raid, and proposed 2024 broadcasting regulations) which create a natural setting to examine the substitutability of foreign information channels.

Methodology
The study employs a novel empirical strategy to identify the causal impact of foreign economic news on local equity markets, addressing the endogeneity inherent in media coverage.

• Data: The analysis utilizes daily data from January 2017 to December 2022. The media variable is constructed using the Global Database of Events, Language, and Tone (GDELT). The primary endogenous variable, media, is the log share of foreign economic news mentioning "India" relative to total global economic news. Financial data (SENSEX index returns, cross-sectional absolute dispersion, conditional variance, and turnover) are sourced from Bloomberg.
• Instrumental Variable (IV): To establish causality, the author constructs a novel instrument based on the calendar dates of public holidays and religious observances in India's Emerging Market (EM) peers (excluding India). The logic relies on editorial substitution: when foreign editors face non-eventful days in other EMs (due to holidays), they are more likely to substitute content by covering India to maintain news volume. This instrument is exogenous to forward-looking financial markets as holiday dates are known in advance and unrelated to economic fundamentals.
• Econometric Model: The author estimates a Proxy-VAR (Vector Autoregressive) model, following Stock and Watson (2012) and Mertens and Ravn (2013). The endogenous vector includes the media proxy, Indian equity returns, cross-sectional absolute dispersion (CSAD), conditional variance (GARCH term), and turnover. The model is identified by instrumenting the foreign news shock, ensuring the instrument correlates with the news shock but is uncorrelated with other structural shocks in the system. Controls include global risk (VIX), S&P 500 returns, and dummies for local holidays and disasters.

Conceptual Framework
The paper develops a theoretical framework where foreign media outlets act as delegated monitors. Unlike models focusing solely on investor attention allocation, this framework emphasizes editorial choices.

• Signal Structure: A news signal is a linear combination of an aggregate (country-specific) fundamental and an idiosyncratic (company-specific) component.
• Learning Mechanism: Investors cannot immediately disentangle these components. However, by reading more articles from diverse outlets (increasing the number of sources, $M$), investors average out the outlet-specific noise, thereby improving their precision in estimating the aggregate fundamental.
• Prediction: Increased foreign news coverage improves the prediction of fundamentals, leading to reduced aggregate market volatility but increased cross-sectional dispersion. This contrasts with herding behavior, which typically results in high volatility and low dispersion.

Key Results
The impulse response functions (IRFs) from the Proxy-VAR reveal the following causal effects of an unexpected increase in foreign economic news mentioning India:

1. Returns: Indian equity returns decrease immediately. This is attributed to the typically negative tone/sentiment embedded in foreign media coverage of EMs.
2. Volatility: Conditional volatility (market-wide risk) decreases on impact and remains lower.
3. Dispersion: Cross-sectional absolute dispersion (CSAD) increases over the subsequent days.
4. Turnover: Trading volume shows no statistically significant reaction, suggesting the price movement is not driven by capital flight or a rush of uninformed investors.

These findings are robust across specifications and placebo tests (showing the instrument does not predict global risk variables like the VIX or S&P 500). Further analysis indicates that high-beta stocks experience larger drops, while government-owned stocks show muted reactions, supporting the interpretation of risk repricing rather than indiscriminate capital flight. Additionally, Google Trends data provides corroborative evidence that increased media coverage is associated with increased search interest in "India" by US readers, supporting the information transmission channel, though the author notes this should be viewed as corroborative rather than definitive evidence.

Significance and Contributions
The paper makes two primary contributions to the literature:

1. Conceptual: It offers a framework explaining how foreign media acts as gatekeepers in EMs. By blending aggregate and idiosyncratic information, foreign coverage allows investors to better estimate fundamentals, thereby reducing aggregate volatility while increasing cross-sectional dispersion. This challenges the narrative that foreign media exposure inevitably leads to coordinated herding and market inefficiencies.
2. Empirical: It introduces a novel instrumental variable for media coverage based on the substitution effect of editorial choices across EM peers. This approach successfully isolates the causal impact of foreign news on local markets.

Implications
The study suggests that for emerging markets, foreign media can serve as a stabilizing force by improving information efficiency and reducing volatility, provided the news flow is not driven by social media frenzies (which tend to increase correlation and volatility). For foreign investors, relying on diverse traditional media sources may mitigate volatility spikes compared to correlated narratives found in social media. The findings imply that in environments where local media is compromised, foreign media can effectively substitute for distorted local information, facilitating better capital allocation.