TY - JOUR T1 - Gains and Losses Revisited: Skill Detection and Similarity Assessment JF - The Journal of Financial Data Science DO - 10.3905/jfds.2022.1.105 SP - jfds.2022.1.105 AU - Sid Browne Y1 - 2022/09/09 UR - https://pm-research.com/content/early/2022/09/10/jfds.2022.1.105.abstract N2 - The article develops an analytical framework that enables investors who use gain- and loss-based performance measures to evaluate and compare investment strategies or managers and to do so in a more precise manner that accounts for statistical uncertainty and sampling error. In particular, the article develops tests for detection of timing skill and sizing skill for individual strategies as well as tests to compare similarity across competing strategies or between different periods (e.g., backtest and live). Some of these tests are exact and therefore relevant for small samples or track records. The article illustrates the methodology throughout by applying these results to a suite of systematic strategy indexes. ER -