Local GMM Estimation for Nonparametric Time‐Varying Coefficient Moment Condition Models
针对非参数时变系数矩条件模型，提出局部连续更新GMM估计量，推导一致收敛率和渐近正态性，讨论带宽选择、置信区间和过度识别检验，并通过蒙特卡洛模拟和资产定价应用验证有限样本表现。
ABSTRACT We develop a local continuously updated GMM estimator for nonparametric time‐varying coefficient moment condition models. The uniform consistency rate and the pointwise asymptotic normality of the proposed estimator are derived. Implementation issues regarding bandwidth selection, construction of pointwise confidence intervals, and testing for overidentifying restrictions are discussed. The finite sample performance of the proposed estimator and test statistic are investigated through a Monte Carlo study and an empirical application on asset pricing models with stochastic discount factor (SDF) representation.