Why does a CFA benchmark need to be specified before the performance period?

Because benchmark timing protects the evaluation from hindsight. If the benchmark is chosen after the result is known, the evaluator can unintentionally select the comparison that makes the manager look best or worst.

A valid benchmark should be documented before the evaluation period so the manager knows the assignment and the client knows the yardstick. Then excess return, tracking error, and information ratio have a stable reference point.

Example: if a small-cap value manager is assigned the North Harbor Small Value Index on January 1, year-end excess return is meaningful. If the manager waits until December 31 and chooses whichever index is easiest to beat, the comparison is not a clean skill test.