Liquidity Risk in Macro-Political Prediction Markets

A Polymarket Case Study

Advanced Research Project High School Elite Scholars Summer 2025

Research Team

Chloe Shao, Nicky Wang
Supervisors: Shukun Shi, Chanjuan Pan
Teaching Assistant: Katherine Liang

Duration 6 weeks
Department VC Investment Research Team
Program Track Camellia Elite Scholars Program High School Track
Research Scope 4 Macro Events
Analysis Period July - August 2025

Abstract

This comprehensive study examines liquidity risk and market dynamics in online prediction markets through detailed analysis of four high-impact macro events on Polymarket. Using both hourly and bi-daily automated data collection over periods, we tracked key liquidity metrics across geopolitical, technology, and policy-driven markets to understand how external information flows influence trading patterns and market efficiency. Our research reveals distinct liquidity profiles across event types: geopolitical markets exhibit high volatility with wide bid-ask spreads in response to unverified news, while specialized policy-focused markets demonstrate more stable pricing with informed participation. High-probability events show greater volatility than low-probability markets, with price fluctuations strongly correlating with information shocks, bid-ask spreads, and trading volume spikes. These findings demonstrate that prediction markets serve dual roles as investment tools and real-time information aggregators, with liquidity dynamics significantly influenced by the nature, credibility, and verification status of external news sources. Markets with broader public interest exhibit higher trading volumes but greater volatility, while specialized markets show superior price efficiency and market sophistication.

Keywords

Prediction markets Polymarket Liquidity risk analysis Market volatility patterns Information flow dynamics Geopolitical forecasting AI competition Public policy Tariff rates

Research Events

Geopolitical Events
Israel x Hamas Ceasefire Before August
Selected for its high volatility and emotional trading responses to geopolitical developments. Demonstrated sharp price movements correlating with ceasefire proposals and diplomatic developments.

Technology Competition
Best AI Model End of July
Chosen to analyze stable probability assessment in technology markets. Exhibited consistent Google dominance with 95.7% probability and low volatility across the analysis period.

Federal Reserve Policy
Jerome Powell Removal 2025
Selected due to ongoing controversy and public debate surrounding Federal Reserve leadership. Demonstrated higher trading volume but greater volatility in response to unverified rumors.

Trade Policy
China Tariff Rates August 15
Chosen due to researcher's unique perspective bridging both sides of U.S.-China trade policy. Showed tighter spreads and more informed participation compared to public-interest markets.

Methodology

Data Collection Process

Automated Python scripts interfacing with Polymarket's Gamma API for consistent time-series observations
Collection Framework
  • Chloe Shao: Hourly data collection over 2-week period (152 data points)
  • Nicky Wang: Bi-daily collection over 2-week period
  • Data stored in CSV format and processed using Pandas
  • Cross-verification against Polymarket webpage displays
  • Data cleaning including duplicate removal and timestamp standardization

Metrics Tracked

Six key liquidity and market efficiency indicators monitored across all events
Core Variables
  • Yes/No share prices - probability-based market pricing
  • Bid-ask spread - liquidity efficiency measurement
  • Total trading volume - cumulative market activity
  • 24-hour trading volume - short-term activity patterns
  • Liquidity pool size - market depth and stability
  • Price volatility - market response to information shocks

Prediction Market Event Analysis

Israel x Hamas Ceasefire Before August
High volatility geopolitical market responding to diplomatic developments
Mean price 0.124
📈 Trading volume $9.3M average
Key Findings
  • Sharp price movements correlating with external events
  • High standard deviation (0.1491) indicating significant volatility
  • Strong correlation between price spikes and ceasefire proposals
  • External information flow significantly influences market dynamics
Best AI Model End of July
Stable technology competition market with clear market leader
Google 95.7% probability
📈 Low bid-ask spread 0.55%
Key Findings
  • Consistent Google dominance throughout analysis period
  • Low volatility with efficient price discovery mechanisms
  • Response to verified corporate earnings announcements
  • Clear differentiation between market leader and competitors
Jerome Powell Removal 2025
Public-interest policy market with high emotional response patterns
Probability dropped 22% → 11%
📈 $2.7M liquidity pool
Key Findings
  • Higher trading volume but greater volatility compared to specialized markets
  • Strong response to unverified rumors and speculation
  • Probability decline following rumor debunking
  • Wider spreads during periods of uncertainty
China Tariff Rates August 15
Specialized trade policy market with informed participant behavior
Average 1.2% spread
📈 Stable 25-40% probability trend
Key Findings
  • Consistently tighter spreads indicating higher market efficiency
  • Gradual price adjustments responding to verified policy announcements
  • Stable liquidity depth with sophisticated participant base
  • 25-40% range emerged as consensus through steady upward trend

Conclusions

Market Type Differentiation

Geopolitical and public-interest events generate high-volume, high-volatility markets with emotional trading responses, while specialized policy-focused markets demonstrate sophisticated participant behavior with efficient pricing mechanisms. Technology competition markets show stable probability assessment with clear market leaders emerging early and maintaining dominance.

Information Flow Impact

External information credibility significantly influences liquidity dynamics. Unverified rumors cause sharp price movements in public-interest markets, while verified policy announcements generate measured, sustained adjustments in specialized markets. This dual role positions prediction markets as both speculation venues and information processing systems.

Investment and Research Applications

Prediction markets serve as valuable real-time sentiment indicators for investors and policymakers. Their liquidity patterns reveal market sophistication levels and information processing efficiency, offering insights for risk assessment and event probability evaluation across diverse sectors. Market popularity, news cycles, and proximity to resolution directly influence liquidity levels and trading patterns.