Market Making Mechanisms and Liquidity Dynamics in Blockchain-Based Prediction Markets
Introduction
Prediction markets represent a sophisticated financial mechanism that harnesses collective intelligence to forecast future events by enabling participants to trade tokenized shares representing different outcomes. In these markets, token prices serve as real-time probability indicators—for instance, a token trading at 0.75 implies a 75% market-assessed probability of the underlying event occurring. The accuracy and efficiency of these probability estimates depend critically on market liquidity and the mechanisms facilitating price discovery.
Contemporary blockchain-based prediction markets employ two distinct trading architectures, each with unique liquidity provision dynamics and economic incentives.
Automated Market Makers (AMMs) utilize algorithmic pricing models based on constant function formulas, where liquidity providers (LPs) deposit paired assets into pooled reserves. While AMMs offer continuous liquidity and low transaction fees for traders, LPs face exposure to impermanent loss—the opportunity cost arising from price divergence between holding assets in the pool versus holding them separately. Conversely,
Central Limit Order Books (CLOBs) facilitate trading through order matching systems where market makers actively quote bid-ask spreads. Although CLOBs enable sophisticated trading strategies and potential arbitrage profits for market makers, they typically impose higher transaction costs on price takers and require continuous active management.
The resolution of prediction markets relies on
oracle mechanisms that provide authoritative outcome determination, creating unique temporal dynamics where market activity intensifies as resolution approaches. This temporal component, combined with the binary or categorical nature of most prediction market outcomes, creates distinct liquidity patterns not observed in traditional financial markets.
Research Objectives
This study investigates the economic viability and optimization strategies for liquidity provision in blockchain-based prediction markets through three interconnected research questions:
RQ1: Profitability Analysis of AMM Liquidity Provision
What factors determine the profitability of AMM liquidity provision on Polymarket, and how do market dynamics influence LP returns?
Utilizing comprehensive historical transaction data from Polymarket (covering the period through 2022), we examine whether providing liquidity in prediction market AMMs generates positive risk-adjusted returns.
RQ2: Strategic Optimization for AMM Liquidity Providers
How can liquidity providers optimize their allocation strategies to maximize risk-adjusted returns in prediction market AMMs?
We analyze behavioral patterns and strategic frameworks employed by successful LPs, examining:
- Temporal allocation strategies including entry/exit timing relative to event resolution, market sentiment shifts, and volume patterns
- Dynamic liquidity management strategies such as position sizing, diversification across event categories, and rebalancing frequencies
RQ3: Market Making Profitability in Central Limit Order Books
What determines the profitability of active market making strategies in CLOB-based prediction markets?
We investigate the economic drivers of market making performance through m
aker-taker fee optimization analyzing how fee structures influence optimal bid-ask spread strategies and position management.
Inventory management strategies including optimal position sizing, hedging techniques, and risk controls for market makers.
Data Sources
Our empirical analysis leverages multiple data streams including:
- On-chain transaction data from Polygon network covering all Polymarket AMM interactions
- Historical data through subgraphs
- Order book snapshots and trade-by-trade data for CLOB analysis
- Oracle resolution data linking market outcomes to actual event results