Asset pricing when traders produce payoff through hard computational search

Authors: Hassan Andrabi, Peter Bossaerts, Konstantinos Ioannidis, Robert Woods, Nitin Yadav

Stage: Draft in preparation

Abstract: We study asset markets in which traders generate the payoff themselves. Each trader privately searches for solutions to a computationally hard (NP-hard) problem, and the asset pays the value of the best solution anyone finds during trading. In standard asset pricing models, by contrast, the payoff is either exogenous or a future equilibrium price. Theory predicts that payoffs will fall short of the optimum but that prices will be efficient. In an experiment, payoffs exceed this prediction: in every market at least one participant finds the optimal solution, although fewer participants do so as instances get harder. Prices converge to the payoff, without the bubbles or mirages common in information-aggregation experiments.

Funding: Society for Experimental Finance ($4,000)

Presented at: 2^nd Summer School in Cognitive Foundations of Decision-Making (Ghent, 2025), 2^nd Annual Conference of UK-based Behavioural and Experimental Economists (Sheffield, 2025), 15^th Annual Society for Experimental Finance Conference (Maastricht, 2025), Tepper School of Business Economics Seminar (Pittsburgh, 2025), 14^th Annual Society for Experimental Finance Conference (Stavanger, 2024)