A Spatial Model of Crop Yield Distributions: Implications for Crop Insurance pricing and Systemic Risk Assessment

刘笑天，华中农业大学副教授，普渡大学经济学博士。主要研究领域为农业保险、粮食安全、空间计量经济学。其研究成果发表于S patial Economic Analysis ，E conometrics 等国际期刊。主持国家自然科学基金青年项目、气象风险与保险开放研究基金等基金项目。

Abstract

This study introduces a novel spatial autoregressive model equipped with heterogeneous coefficients and latent group structures for crop yield insurance pricing and systemic risk assessment. A nonparametric Bayesian estimation method is employed to estimate the model. Monte Carlo simulations show its good finite-sample performance.

The newly developed model offers at least two key advantages. First, it integrates a spatially lagged dependent variable to address spatial correlations in crop yields, enhancing the accuracy of distribution estimates. This feature allows for the direct computation of joint distributions, presenting an effective alternative to traditional copula models for systemic risk evaluation. Secondly, the model's latent group structure adeptly captures the variability in growing conditions across different years, thereby offering increased flexibility for modeling technological changes.

We apply the proposed model to measure risks of simultaneous global breadbasket failure and calculate area-yield crop insurance premiums. An out-of-sample game, simulating a scenario between a private insurance company and the federal agency, indicates that the proposed method is more efficient at estimating premium rates compared with current RMA method.

主持人：付博文

湖南大学经济管理研究中心助理教授

悉尼科技大学经济学博士

时间：2023年11月17日下午4:15-5:30

水教211