A new study by Utpal Bhattacharya, Tse-Chun Lin, and Janghoon Shon finds that Hong Kong’s 2020 National Security Law led local financial analysts to self-censor their reports, particularly when covering poorly performing state-owned enterprises.
Following the introduction of Hong Kong’s National Security Law (NSL) in June 2020, local financial analysts began self-censoring their reports, particularly when covering poorly performing state-owned enterprises, according to new research by Utpal Bhattacharya, Tse-Chun Lin, and Janghoon Shon. The findings suggest that constraints on political speech can have far-reaching effects on financial discourse and market efficiency.
The new study examines how the NSL’s enactment influenced the behavior of equity analysts covering major companies listed on the Hong Kong Stock Exchange. Their analysis reveals that local analysts began providing more optimistic forecasts, using vaguer language, and taking longer to issue reports about underperforming firms after the law took effect—patterns that were especially pronounced when covering Chinese central state-owned enterprises (SOEs). SOE’s are legal entities that undertake commercial activities on behalf of the Chinese central government (as opposed to provincial or city governments).
The Law’s impact on financial speech
The NSL, which came into effect on June 30, 2020, prohibits acts of secession, subversion, terrorism, and collusion with foreign forces. While the law’s primary focus is political speech, the researchers find evidence that its effects spilled over into financial analysis, likely due to uncertainty about how broadly the law might be interpreted.
Using data from 2018 to 2022, the study examines over 6,000 analyst reports covering 38 major companies consistently listed in Hong Kong’s Hang Seng Index. The researchers compared the behavior of local analysts (identified by Chinese family names) with foreign analysts before and after the NSL’s implementation, focusing particularly on how they covered firms during periods of poor performance.
To identify potential self-censorship, the authors examined three key metrics: forecast errors in earnings predictions, the use of weak modal words in reports (such as “could,” “might,” and “perhaps”), and the time lag between earnings announcements and analyst reports. They also analyzed market reactions to analyst recommendations to assess whether investors adjusted their behavior in response to potential self-censorship.
The study controlled for various factors that could influence results, including changes in the teams of analysts that issued the reports, drastic changes to firm stock prices, and seasonal variation in firm performance.
Three key changes in analyst behavior
The study documents changes in analyst behavior across three dimensions:
First, local analysts’ earnings forecasts showed an upward bias after the NSL when covering underperforming firms. Compared to foreign analysts, local analysts’ forecast errors were 0.47 to 0.63 standard deviations higher for poorly performing companies after the law’s enactment. This effect was even stronger for central SOEs, where the difference in forecast errors between local and foreign analysts was 1.39 standard deviations, compared to just 0.26 standard deviations for non-SOEs.
Second, when writing about central SOEs with poor performance, local analysts doubled their use of weak modal words compared to foreign analysts after the NSL. This increase in ambiguous language suggests analysts may have been attempting to hedge their statements when discussing negative developments.
Third, local analysts took longer to issue reports following earnings announcements from poorly performing central SOEs after the NSL. The study found that local analysts’ response time was about 26 days longer than foreign analysts when covering underperforming central SOEs post-NSL implementation.
Market recognition of self-censorship
The market appears to have recognized and adjusted for this self-censorship. After the NSL’s implementation, stock price reactions to local analysts’ buy and neutral recommendations compared to sell recommendations decreased significantly, particularly for central SOEs. The market responded 5.2 percentage points less positively to buy recommendations compared to sell recommendations from local analysts compared to foreign analysts for central SOEs, while the difference was only 2.6 percentage points for non-SOEs.
This muted market response suggests investors discounted the information content of positive recommendations from local analysts, recognizing potential self-censorship in their analysis.
Broader implications
The study’s findings raise important questions about information quality in financial markets operating under political constraints. While the NSL doesn’t explicitly restrict financial analysis, uncertainty about its scope appears to have led analysts to err on the side of caution, particularly when discussing state-owned enterprises.
This self-censorship could have meaningful implications for market efficiency. Accurate financial analysis plays a crucial role in price discovery and capital allocation. If analysts feel constrained in their ability to provide objective assessments, it may become more difficult for investors to make well-informed decisions.
The researchers note that their findings expand upon previous research showing how government influence on media can affect various forms of public discourse. While prior studies have documented direct effects of political control on media coverage, this research suggests that even uncertainty about potential regulatory consequences can lead to self-censorship in adjacent domains like financial analysis.
The study also builds on existing literature about analysts’ tendency to moderate negative opinions due to career concerns and business relationships. They may fear that negative recommendations or forecasts could lead to a direct “retaliation” from the companies they cover, such as termination of investment banking business relations and other adverse effects on their career. The introduction of political and legal uncertainties appears to have added another layer of complexity to these already existing pressures toward positive bias.
These findings contribute to our understanding of how political speech restrictions can affect financial markets and highlight the potential for regulatory uncertainty to influence behavior even in domains not explicitly targeted by regulation. As markets globally grapple with various forms of political intervention, this research suggests that the indirect effects of such interventions on market participants’ behavior may be as important to consider as their direct effects.
Note: Readers may contact the authors for a copy of their paper.
Author Disclosure: the author reports no conflicts of interest. You can read our disclosure policy here.
Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.
