Master's Thesis Parshant Singh
Market Making Mechanisms and Liquidity Dynamics in Blockchain-Based Prediction Markets
Abstract
Prediction markets have increasingly emerged as alternatives to traditional polling methods. Decentralised designs, in particular, reduce reliance on central operators and enable more transparent forecasting. A persistent challenge, however, lies in liquidity provision. Automated Market Makers (AMMs), while popular in decentralised finance, suffer from impermanent loss, an issue that is especially severe in prediction markets, since one of the outcome tokens becomes worthless once the event resolves. This study focuses on Polymarket, a leading decentralised prediction market built on Polygon, and compares two market designs it has employed: AMM-based and Central Limit Order Book (CLOB)-based systems. Our analysis shows that liquidity provision in the CLOB-based design is significantly more profitable than in the AMM-based design, where most potential profits are eroded by realised impermanent loss. We also find that LP strategies differ across the two systems. In AMMs, providers often use Just-in-Time (JiT) liquidity and remove liquidity before market resolution to reduce exposure to impermanent loss. In CLOBs, LPs dynamically adjust bid–ask quotes to manage inventory risk while maximising platform rewards. These results highlight how market design directly impacts LP behaviour and profitability, and offer insights for the development of more sustainable prediction markets.
Research Questions
- What factors determine the profitability of liquidity provision in AMMs on Polymarket prediction markets?
- How can liquidity providers optimise their strategies to improve profitability in prediction market AMMs?
- What factors determine the profitability of liquidity provision in CLOB-based prediction markets?
| Attribute | Value |
|---|---|
| Title (de) | Market-Making Mechanismen und Liquiditätsdynamiken in blockchain-basierten Prognosemärkten |
| Title (en) | Market Making Mechanisms and Liquidity Dynamics in Blockchain-Based Prediction Markets |
| Project | |
| Type | Master's Thesis |
| Status | completed |
| Student | Parshant Singh |
| Advisor | Jonas Gebele |
| Supervisor | Prof. Dr. Florian Matthes |
| Start Date | 08.04.2025 |
| Sebis Contributor Agreement signed on | |
| Checklist filled | Yes |
| Submission date | 03.09.2025 |