No CrossRef data available.
Published online by Cambridge University Press: 09 November 2020
We propose a multiplex interdependent durations model and study its empirical content. The model considers an empirical stopping game of multiple agents making optimal timing decisions with incomplete information. We characterize the unique Bayesian Nash equilibrium of the stopping game in a system of simultaneous equations involving the conditional distribution of each duration with a moderate strategic interaction condition. The system of nonlinear simultaneous equations allows us to obtain constructive identification results of the interaction effects and other nonparametric model primitives. We propose two consistent semiparametric estimation methods based on different parameterizations of modeling components with right-censored duration data.
This paper benefited greatly from insightful comments and suggestions provided by co-editor, Sokbae Lee, and two reviewers. We are grateful to Jaap Abbring, Jeffrey Campbell, Yingyao Hu, Qi Li, Xiaodong Liu, Essie Maasoumi, Áureo de Paula, Yuya Sasaki, Tiemen Woutersen, and Haiqing Xu for many thoughtful discussions. We also thank the audience at Tilburg University, Johns Hopkins University, University of Colorado, Boulder, and the ES 2017 European Meeting. All errors remaining are our own.