Effects of Market Information Costs on Efficiency and Concentration in AI Model Markets with Arbitrage

Determine how positive costs of acquiring market information in AI model marketplaces affect whether computational arbitrage improves market efficiency or instead shifts market concentration from model providers toward oligopolistic arbitrage intermediaries.

Background

The analysis assumes perfect market information, which simplifies consumer and arbitrageur behavior by presuming full knowledge of prices and performance.

The authors explicitly raise the unresolved question of how the cost of market information could affect market outcomes, including whether arbitrage enhances efficiency or leads to concentration among intermediaries.