ECONOMETRIC ANALYSIS OF CONTINUOUS TIME MODELS: A SURVEY OF PETER PHILLIPS’S WORK AND SOME NEW RESULTS
综述彼得·菲利普斯在连续时间计量经济学领域40年的贡献，涵盖从单方程到方程组、从渐近理论到有限样本、从参数到非参数模型等问题，并基于其早期工作比较三种渐近分布的表现。
Econometric analysis of continuous time models has drawn the attention of Peter Phillips for 40 years, resulting in many important publications by him. In these publications he has dealt with a wide range of continuous time models and the associated econometric problems. He has investigated problems from univariate equations to systems of equations, from asymptotic theory to finite sample issues, from parametric models to nonparametric models, from identification problems to estimation and inference problems, and from stationary models to nonstationary and nearly nonstationary models. This paper provides an overview of Peter Phillips’ contributions in the continuous time econometrics literature. We review the problems that have been tackled by him, outline the main techniques suggested by him, and discuss the main results obtained by him. Based on his early work, we compare the performance of three asymptotic distributions in a simple setup. Results indicate that the in-fill asymptotics significantly outperforms the long-span asymptotics and the double asymptotics.