Export Reviews, Discussions, Author Feedback and Meta-Reviews

Neural Information Processing Systems 

Summary of the paper: The paper aims to address the question of how to combine beliefs to a central judgement in the case that the beliefs held by agents have an information asymmetry in each agents perspective on the future outcomes of particular stochastic events. In particular the paper particularly focuses on the concept of Hanson 2003 known as the prediction market based on a logarithmic market scoring rule. The market scoring rule framework is sometimes also known as a market studied in the context of cost function based market makers. The motivation provided for the MSR setting involves financial markets in which market makers or dedicated specialists are provided incentives, typically in the form of rebates to take trades on both sides of the book ie. the place limit and market orders in such a way that they create liquidity in the market to facilitate buying and selling of an asset. In this context it is argued that a MSR effectively acts as a cost function-based market maker always willing to take the other side of a trade with any willing buyer or seller, and re-adjusting its quoted price after every transaction.