City Research Online

Abstract

This thesis investigates the role of information and algorithms in shaping competition, strategy, and welfare outcomes in digital markets. Across three chapters, it examines how algorithmic decision-making, data advantages, and platform dual roles influence firm behavior and consumer outcomes.

The first strand explores whether large language models (LLMs) can learn predatory strategies in dynamic environments in which an incumbent faces repeated entry threat. Using OpenAI’s GPT-4.1 as decision-making agents, we find that LLMs learn to predate when both predation and accommodation are theoretically viable, and adopt aggressive strategies when only accommodation is theoretically viable. Further, profit optimization is limited, highlighting both strategic learning and its limitations. These results reveal that LLMs are capable of executing complex exclusionary strategies.

The second strand considers a model where two competitors located at opposite ends of the Hotelling line compete for the unit mass of consumers. Consumer preferences vary across two dimensions: horizontal (brand preference) and vertical (quality preference). We compare two scenarios :(1) competitors remain uninformed about the horizontal dimension but possess perfect information about the vertical dimension, which enables them to set prices conditional on this information; (2) competitors are uninformed about either dimension of consumer preferences and set a uniform price. The analysis reveals that under some conditions consumers benefit from personalized pricing.

The third strand examines the incentives of a hybrid platform, which acts both as a marketplace provider and as a competitor through its own retail arm, to truthfully share demand information with a third-party seller. The platform enjoys an informational advantage about the state of demand, which it always shares with its retail arm. We analyze two settings: (i) No hosting fee i.e, the seller pays no commission and (ii) Platform charges a hosting fee i.e, commission a percentage of seller revenue. In both cases, we find that the incentives of the platform and the seller diverge, leading to an uninformative equilibrium. While commissions affect prices and profit levels, they do not alter the qualitative nature of information sharing. These results highlight the limits of informational advantage in hybrid platform markets.

Publication Type:	Thesis (Doctoral)
Subjects:	H Social Sciences > HF Commerce Q Science > QA Mathematics > QA75 Electronic computers. Computer science Z Bibliography. Library Science. Information Resources > ZA Information resources Z Bibliography. Library Science. Information Resources > ZA Information resources > ZA4050 Electronic information resources
Departments:	School of Policy & Global Affairs > Department of Economics School of Policy & Global Affairs > School of Policy & Global Affairs Doctoral Theses Doctoral Theses

Export

Downloads