New and Old Sorts: Implications for Optimal Portfolios