Abstract
Shedding Light on the Role of Sample Sizes and Splitting Proportions in Out-of-Sample Tests: A Monte Carlo Cross-Validation Approach