A new research paper, recently disseminated on arXiv, delves into the fundamental question of how information is aggregated in prediction markets, specifically comparing Central Limit Order Book (CLOB) and Automated Market Maker (AMM) designs. The preprint, titled "Who Aggregates Information? Screening, Rent, and the Coexistence of CLOB and AMM Prediction Markets," was authored by Chengqi Zang, Gabriel Andrade, and Tomoyuki Nakajima. Published on August 31, 2026, the study offers insights into the conditions under which these distinct market mechanisms might coexist and their implications for market efficiency and participant incentives.
Market Design and Information Aggregation
The paper explores the differing ways CLOB and AMM structures facilitate information aggregation. CLOB markets, common in traditional finance, rely on explicit bids and offers from traders, allowing for direct price discovery through order matching. AMMs, prevalent in decentralized finance, use mathematical functions to determine prices and liquidity, often without direct counterparty matching. The researchers investigate how these structural differences influence the ability of markets to incorporate new information and reflect accurate probabilities of future events.
Screening and Rent Extraction
A key focus of the study is the concept of "screening"—how market designs enable or hinder informed traders from profiting from their private information—and "rent extraction," referring to the profits generated by market makers or liquidity providers. The paper likely models how the presence of informed traders interacts with the specific mechanics of CLOBs versus AMMs, affecting the profitability of providing liquidity and the overall efficiency of price discovery. Understanding these dynamics is crucial for designing robust and fair prediction markets.
Coexistence Dynamics
The research also examines the conditions necessary for the coexistence of both CLOB and AMM prediction markets. This aspect of the study is particularly relevant as the prediction market landscape sees a mix of centralized platforms often employing CLOBs and decentralized, onchain markets frequently utilizing AMMs. The authors likely analyze factors such as liquidity provision, transaction costs, and the types of information available to traders to determine scenarios where both models can thrive simultaneously, potentially catering to different segments of the market or types of events.
Background on Market Structures
Prediction markets serve as venues for trading contracts whose value is tied to the outcome of real-world events. Their design is critical to their effectiveness as forecasting tools. CLOBs offer transparency and direct interaction between buyers and sellers, while AMMs provide continuous liquidity and are often favored for their simplicity and automation, particularly in decentralized settings. The ongoing academic inquiry into these market designs helps refine our understanding of how best to structure these platforms to maximize their forecasting accuracy and utility.
What to Watch
As a preprint, this paper represents an early stage of academic research, and its findings will likely be subject to further peer review and discussion within the economic and market microstructure communities. Future research may build upon these models to test their hypotheses with empirical data from existing prediction markets or explore hybrid market designs that combine elements of both CLOBs and AMMs. The insights from such studies could inform platform operators and regulators on optimal market structures for enhancing information aggregation and market integrity.
Original announcement: arXiv