ABSTRACT:
In markets such as rental housing, self-storage, equipment leasing, and cloud computing, capacity turns over rather than depletes as customers arrive randomly, occupy units for uncertain durations, and eventually leave. Pricing low may fill a unit today but forgoes the option of renting it at a higher price tomorrow. Firms at full capacity drop out of consumers' choice sets, altering substitution patterns in real time.
We model this setting as a continuous-time Markov process and compute dynamic equilibria under three regimes: firms price based on their own occupancy, on all firms' occupancies, or to jointly maximize profits. Three results emerge. First, sharing occupancy information intensifies competition, as firms compete in each realized state rather than on average. Second, as occupancy rises, it becomes more difficult to absorb diverted rival demand, so firms become worse substitutes for each other, shrinking the gains to collusion. Third, scale pools demand risk, raising occupancy and lowering prices.
Together, these findings identify a structural channel through which algorithmic pricing in capacity-constrained markets generates efficiency gains rather than collusive overcharges. Instead of asking whether trigger strategies can sustain supra-competitive prices, we ask the prior question, whether the prize is large enough to justify the antitrust risk. The results yield practical guidance on which pricing algorithm to adopt and what information to feed it.
No comments:
Post a Comment