Mean Absolute Scaled Error as imputation performance metric

Question

To evaluate different imputation methods with cross-validation I am searching for an appropriate accuracy measure. My cross-validation sample consists of 100 univariate time