I am a PhD candidate at the Toulouse School of Economics.
My main research interests are Industrial Organization, Competition Policy, Computational Methods, and Machine Learning.
In my work, I use structural models and Reinforcement Learning methods to study dynamic aspects of competition.
I am available for interviews during the 2025-2026 Job Market.
Here is my CV.
You can contact me at maxim.sandiumengeiboy@tse-fr.eu.
Consumer Dynamics and Vertical Relations: Coordination and Foreclosure in a U.S. Consumer-Goods Industry (Job Market Paper)
Abstract
Vertical mergers along the supply chain can generate efficiency gains by improving coordination, and anti-competitive harms by disadvantaging rivals. However, they are typically analyzed in a static setting due to their complexity, limiting our ability to assess recent policy concerns about their dynamic consequences. This paper evaluates how demand-induced dynamics reshape the effects of vertical mergers and uses deep reinforcement learning to overcome the associated computational challenges. To do so, I develop a dynamic model in which downstream firms have multiple suppliers and face dynamic demand, and I train neural networks to approximate the Markov perfect equilibrium. I estimate the model using data from a U.S. consumer-goods industry, where habit formation in demand induces strong dynamics. The findings reveal that dynamic considerations magnify the consequences of any competitive disadvantage. This prompts firms to moderate their margins, but also strengthens their incentives to disadvantage non-integrated suppliers. Together, these forces produce a perverse outcome: as demand dynamics strengthen, efficiency gains from integration shrink by up to 35% compared to the static case, and integrated firms reduce prices of integrated products up to 30% less, while foreclosing non-integrated products more severely. Overall, intertemporal linkages dampen the pro-competitive effects of vertical mergers and amplify their anti-competitive risks.
Abstract
This paper shows that intertemporal linkages can affect the capability and incentives to vertically foreclose rivals. Intertemporal linkages introduce two opposing effects. First, they magnify the impact of cost asymmetries (direct effect), potentially enhancing foreclosure concerns. Second, they discipline alternative suppliers (indirect effect), potentially reducing the scope for foreclosure. The overall impact depends on the feasibility of a counter-merger and the strength of the intertemporal linkages. When a counter-merger is not possible or the intertemporal linkages are weak, the scope for foreclosure increases. Conversely, when a counter-merger is feasible and the intertemporal linkages are strong, the scope for foreclosure diminishes or disappears entirely.
Endogenous Vertical Networks (with Ali Yurukoglu)
Platform Annexation (with Patrick Rey)
Product Adoption and State Dependence (with Nicolás Martínez)
Graduate Game Theory: 2021, 2022, 2023
Best Teaching Assistant Award: 2023
Graduate Empirical Industrial Organization: 2023
Competition Policy Workshop: 2023
Undergraduate Econometrics: 2022
Undergraduate Industrial Organization: 2021