AI 中文总结
本文针对高维面板数据，在具有交互固定效应的框架下，将高维线性面板模型扩展为可加非线性结构，推导了小T和大T情形下估计量的收敛速率，并通过蒙特卡洛实验与实证应用验证了理论。
AI 中文摘要
现代经济面板数据集常为高维，包含大量控制变量，其数量甚至可能超过样本量。然而，高维面板计量经济学方法的相关文献相当有限。本文研究具有交互固定效应的高维面板模型，其中回归函数具有可加结构，即每个协变量通过未知的非线性分量函数进入模型。我们在该可加框架下开发估计方法与理论，这大幅扩展了Ruecker等人（2025）关于高维线性情形的前期工作。理论部分推导了小T和大T面板情形下估计量的收敛速率，同时通过全面的蒙特卡洛实验和实证应用对理论进行补充验证。
英文摘要
Modern economic panel data sets are often high-dimensional: they contain information on a wide variety of control variables whose number may even exceed the sample size. Nevertheless, the literature on econometric methods for high-dimensional panels is quite limited. In this paper, we study high-dimensional panel models with interactive fixed effects where the regression function has an additive structure, i.e., each covariate enters the model via an unknown nonlinear component function. We develop estimation methodology and theory in this additive framework which substantially extends previous work on the high-dimensional linear case by Ruecker et al. (2025). In the theoretical part of the paper, we derive the convergence rate of our estimator for both the small-T and the large-T panel case. The theory is complemented by comprehensive Monte Carlo experiments and an empirical application.
