报告人： 兰伟 教授

报告题目： Quantile Social Autoregressive Model

时间： 2026年9月11日 16:00—17:00

地点： 西安交通大学兴庆校区数学楼2-2会议室

报告摘要：
Research on modelling peer effects has predominantly relied on linear-in-means models. However, by averaging responses across peers, these models fail to capture situations where influence stems from extreme peer behaviour. To address this limitation, this paper introduces a quantile-based social norm defined on the empirical distribution of peers' responses and proposes a novel quantile social autoregressive model.

In our framework, peers' responses are aggregated by the quantile of their distribution at an unknown level, allowing the data to identify which segment of the peer group drives individual behaviour. For estimation, we develop a new set of moment conditions using instruments constructed from pseudo responses. To address the nonsmoothness of the quantile social norm, we apply kernel smoothing when constructing these instruments and further smooth the quantile social norm appearing in the residual to facilitate statistical inference.

We establish the existence and uniqueness of the equilibrium, along with model identification conditions. Furthermore, we derive the consistency and asymptotic normality of the proposed estimator. Monte Carlo experiments examine the finite-sample performance of the proposed estimator, and an empirical application illustrates the interpretation of the estimated peer effect and quantile level.

报告人简介：
兰伟，西南财经大学教授、博士生导师，现任统计与数据科学学院副院长、财经数智科学创新实验室主任。主要研究方向为大型网络数据分析、实证资产定价和投资组合优化。主持国家自然科学基金优秀青年科学基金项目、面上项目和多个重点项目子课题。已在 Management Science、Journal of the American Statistical Association、Annals of Statistics、Journal of Econometrics、Journal of Business & Economic Statistics、《经济学季刊》等国内外知名期刊发表论文60余篇。