Website Paper
Fields
Industrial Organization, Econometrics
Job market paper
AI and Demand-Based Option Listing
Single-stock options account for over half of U.S. option contracts, yet 40% of single-stock option contracts listed during August 2021–July 2022 never had any trading volume. This paper studies option listing choice, focusing on Nasdaq's August 2022 adoption of an AI-based system that predicts contract-level market-wide demand and adjusts strike listings on Nasdaq-affiliated options exchanges. Using U.S. options transaction data around this adoption, I find that AI listing optimization increased the number of listed contracts by 11.8% and total lifetime trading volume by 30.4% in a difference-in-differences design. Relative to matched stocks listed on other exchanges, Nasdaq's AI-based option listing has the largest effects for low-volatility stocks, whose prices fluctuate relatively little. For these stocks, Nasdaq adds listings near current stock prices. The AI also reduces inactive listings for contracts with weekly expiration dates. I then develop a stylized strike-listing model in which an exchange maximizes fee revenue from volume, market makers bear inventory risk from quoted contracts, and investors value the utility-maximizing attainable portfolio. The model shows that volume-maximizing strikes differ from those that investors value the most. AI-based listing optimization can raise market activity and reduce stale listings, but volume maximization does not imply greater welfare or market completeness.