Missing Values Handling for Machine Learning Portfolios

Documentation, code, and pseudo data for replicating Chen and McCoy (Forthcoming, Journal of Financial Economics, https://arxiv.org/pdf/2207.13071.pdf), to satisfy the journal's code and data sharing requirements. Instead of this packet, we recommend using the code posted on github (https://github.com/jack-mccoy/missing_data), which contains all previous versions of the code and allows for issue tracking (should it be needed). Code from either source automatically downloads stock return data from WRDS and return predictors from https://sites.google.com/site/chenandrewy/, so the pseudo data is not really necessary. You can also find imputed predictor data at https://sites.google.com/site/chenandrewy/.

本数据集包含用于复现Chen与McCoy（即将发表于《金融经济学杂志（Journal of Financial Economics）》，预印本链接：https://arxiv.org/pdf/2207.13071.pdf）研究的说明文档、代码与模拟数据，以满足该期刊对代码与数据共享的要求。相较于本数据包，我们推荐使用发布于GitHub（https://github.com/jack-mccoy/missing_data）的代码：该仓库包含该代码的所有历史版本，并支持问题追踪（如有需要）。任一来源的代码均可自动从沃顿研究数据服务平台（WRDS）下载股票收益数据，并从https://sites.google.com/site/chenandrewy/下载收益预测因子数据，因此本模拟数据实际上并非必需。用户亦可在https://sites.google.com/site/chenandrewy/获取插补后的预测因子数据。