Tong Xie
Ph.D. Candidate in Operations Management, Chicago Booth
Tong Xie

I am a fourth-year Ph.D. candidate in Operations Management at the University of Chicago Booth School of Business, where I am fortunate to be advised by René Caldentey.

I study markets where algorithms interact: how they learn, when they collude, and how platforms should intervene. My earlier work studies assortment optimization and choice overload.

I hold a B.B.A. in Economics from CUHK(SZ) and an M.Phil. in Industrial Engineering from HKUST, where I was fortunate to be advised by Zizhuo Wang and Ying-Ju Chen.

Google Scholar · CV

I will be on the 2026-27 academic job market.

Research

Publications

  1. Personalized Pricing and Assortment Optimization Under Consumer Choice Models with Local Network Effects
    with Zizhuo Wang
    Operations Research, 2024

Working Papers

  1. Speed of Intervention in Algorithmic Markets: Controlling Collusion and Stability (Job Market Paper)
    with René Caldentey, Martin Haugh

    We study how platform interventions shape markets with algorithmic collusion. The effectiveness of intervention depends not only on how strong it is, but also on how fast it is implemented. Sudden interventions may induce instability when pushing algorithms out of the "basin of attraction," leading the market into persistent oscillation. Our results show that effective market regulation requires careful design of both level and speed.

    • Accepted for presentation at 2026 Marketplace Innovation Workshop
    • Accepted for presentation at the Econometric Society Interdisciplinary Frontiers: Economics and AI+ML Conference 2026
    • Accepted for presentation at 2026 INFORMS Revenue Management and Pricing (RMP) Conference
  2. Intertemporal Demand Allocation for Inventory Control in Online Marketplaces
    with René Caldentey

    Online marketplaces do more than match buyers and sellers—they also allocate orders and offer fulfillment services. We study how a platform can influence sellers' inventory decisions through intertemporal demand allocation, without directly controlling stock. We develop a model where the platform observes total demand and assigns orders across sellers, while sellers choose between fulfill-by-merchant (FBM) and fulfill-by-platform (FBP) and manage inventory accordingly. The key mechanism is informational: allocation affects the predictability of each seller's sales, which in turn determines their safety-stock needs, even when average demand shares are unchanged. Focusing on policies that treat sellers equally, we show that uniform splitting minimizes uncertainty, while higher uncertainty can be implemented with simple rules that limit sellers' ability to infer total demand from their own sales. Overall, demand allocation emerges as a key design lever: by adjusting demand predictability, platforms can shape inventory choices and the adoption of FBP.

  3. Integrating Choice Overload into Assortment Planning
    with René Caldentey, Srikanth Jagabathula

    Choice overload is widely believed to hurt firms. We show that it can be beneficial. Using IRI data, we uncover an inverted U-shaped relationship between assortment size and brand switching, indicating that larger assortments can discourage switching among high-value consumers. We then develop a parsimonious framework that incorporates choice overload into assortment planning, offering insights into how firms can optimize product offerings in the presence of choice overload.

    • Accepted for presentation at 2025 INFORMS Revenue Management and Pricing (RMP) Conference
  4. Bundling and Pricing Decisions for Ancillary Products
    with René Caldentey
    • Accepted for presentation at 2024 INFORMS Revenue Management and Pricing (RMP) Conference

Teaching

Teaching Assistant, Chicago Booth