Learning a ranking rather than a level
Question
A portfolio trades only the highest- and lowest-ranked stocks each day. Why might ranking quality and raw prediction MSE lead to different model choices?
Short answer
MSE penalises level errors across all observations, whereas a ranking policy depends on relative ordering and especially the selected tails. Compare cross-sectional rank metrics and realised net portfolio outcomes, while checking calibration for sizing. Ranking objectives can align better with selection but may ignore payoff magnitude and costs. Choose the objective and portfolio rule inside development data and assess them jointly out of sample.
Original question and answer written by SelfMadeQuant.com.
See the hint and the full worked solution in the question bank (free account).
Open the worked solutionLessons behind this question
Start with a free lesson: Expectation, Variance and Covariance is part of a free topic. Create a free account, no card needed, to read it in full with worked solutions.
Read it free