Abstract We construct a model to examine the time-varying ambiguity of investors. When ambiguity occurs concerning recent news, long (short) position investors who are averse to ambiguity reduce (increase) their holdings, resulting in price drops (rises). We empirically analyze how the “two sessions,” a significant event with high policy ambiguity in China, affect the financial market. Our findings suggest that institutional investors mainly sell their holdings between 15 and 5 days before the meetings. Furthermore, the delay in the “two sessions” in 2020 suggests that these selloffs are driven by ambiguity aversion rather than new information.
Introduction
In financial markets, a critical distinction exists between risk—where future outcomes are unknown but their probability distributions are understood—and ambiguity, where even the underlying probabilities are unknown. While standard models assume rational reactions to information, they often overlook how investors behave when they know significant news is coming but cannot quantify its nature.
This paper investigates "time-varying ambiguity" by examining China's annual "two sessions" (Lianghui). These meetings of the National People's Congress (NPC) and the Chinese People's Political Consultative Conference (CPPCC) are pivotal for national policy. Because their dates are announced well in advance and they historically trigger major macroeconomic shifts, they create a period of intense policy ambiguity. We explore whether institutional investors preemptively adjust their portfolios to avoid this ambiguity and how such behavior impacts market prices.
Theoretical Framework: Modeling Ambiguity Aversion
The study develops a theoretical model featuring two primary market participants: ambiguity-averse institutional investors and retail investors (noise traders). Institutional investors are modeled using a max-min utility function, which implies they make decisions to minimize potential worst-case scenarios when faced with ambiguous signals.
As illustrated in the model's demand curves, as an event approaches and ambiguity regarding the outcome increases, these investors reduce their optimal holdings to mitigate exposure to the unknown. This collective reduction in demand from sophisticated players creates downward pressure on stock prices, particularly for those with high institutional concentration.
Empirical Strategy and Data
To test the model's predictions, the researchers analyzed 2,973 A-share stocks listed on the Shanghai and Shenzhen Stock Exchanges from 2002 to 2022. The primary metric for grouping was the proportion of institutional ownership, sourced from the Wind database.
The methodology involved:
- Portfolio Sorting: Stocks were sorted into deciles based on institutional ownership ratios.
- Event Window Analysis: Cumulative abnormal returns (CAR) were calculated for various windows (e.g., 15, 10, and 5 days) prior to the "two sessions."
- Regression Analysis: Panel regressions controlled for firm-specific factors like size, P/E ratio, leverage, and trading volume to isolate the effect of institutional ownership on pre-event returns.
Key Findings: The Pre-Event Selloff
The empirical results strongly support the ambiguity aversion hypothesis. The data reveals a distinct "ambiguity premium" where institutional investors are willing to forgo potential gains to avoid the uncertainty of political announcements.
- Timing of the Selloff: The most significant selling pressure occurs between 15 and 5 days before the meetings. By the final five days leading up to the event, most portfolio rebalancing is complete, and price pressure stabilizes.
- Institutional Impact: Stocks in the top decile of institutional ownership significantly underperformed those in the bottom decile (predominantly retail-held) during the 15-to-6-day pre-event window. Regression coefficients for the institutional ownership ratio were consistently negative and statistically significant.
The 2020 Natural Experiment
A unique opportunity to validate these findings arose in 2020 when the COVID-19 pandemic forced the "two sessions" to be delayed from March 3 to May 22. This delay allowed researchers to distinguish between information leakage and ambiguity aversion.
If the selloff were driven by insider information about specific policies, it would have occurred only before the actual May meetings. However, the market exhibited the exact same institutional selloff behavior ahead of the originally scheduled March date. Since no policies were finalized or even discussed in March, this confirms that the selloff is a psychological reaction to the anticipation of ambiguity, independent of actual fundamental news.
Conclusion and Implications
This research highlights how the scheduling of major political events can inadvertently induce market volatility. While institutional investors are often viewed as the "smart money," their aversion to ambiguity leads to predictable selloffs that impact market stability.
Future Directions: * Trading Strategies: The predictable nature of this selloff suggests potential for long-short strategies—shorting high-institutional-ownership stocks and going long on retail-heavy stocks in the weeks preceding known political events. * Policy Recommendations: Regulators should consider how the signaling and timing of major announcements affect investor behavior. Reducing the "ambiguity window" through clearer communication could help mitigate unnecessary pre-event volatility.
The study's reliance on quarterly institutional data remains a limitation, and future research using higher-frequency ownership data could further refine these insights.
Summary updated