Discussion paper

DP18476 Buyer-Optimal Algorithmic Consumption

We analyze a bilateral trade model in which the buyer's value for the product and the seller's costs are uncertain, the seller chooses the product price, and the product is recommended by an algorithm based on its value and price. We characterize an algorithm that maximizes the buyer's expected payoff and show that the optimal algorithm underrecommends the product at high prices and overrecommends at low prices. Higher algorithm precision increases the maximal equilibrium price and may increase prices across all of the seller's costs, whereas informing the seller about the buyer's value results in a mean-preserving spread of equilibrium prices and a mean-preserving contraction of the buyer's payoff.

£6.00
Citation

Ichihashi, S and A Smolin (2023), ‘DP18476 Buyer-Optimal Algorithmic Consumption‘, CEPR Discussion Paper No. 18476. CEPR Press, Paris & London. https://cepr.org/publications/dp18476