Research / Theme
Manager Selection: Skill vs Luck
Evaluating investment managers is one of the most persistent challenges in institutional investing. Strong historical performance is often interpreted as evidence of skill, yet outcomes are shaped by market conditions, factor exposure, and randomness as much as by decision quality.
Many selection processes rely heavily on backward-looking metrics. While necessary, these measures are insufficient for understanding whether performance reflects durable skill or favorable circumstances. A manager who outperformed during expansion may have been long beta in a bull market. A manager who underperformed during contraction may have preserved capital better than the benchmark suggests.
The regime problem is central to manager evaluation. Performance divorced from regime context is nearly uninterpretable. The relevant question is not what a manager returned — it is what they returned in the environments that test their stated philosophy.
This research theme examines the structural difficulty of separating skill from luck in manager evaluation. It focuses on attribution, consistency, decision-making processes, and the alignment between stated philosophy and observed behavior across different market environments.
MARY applies regime-aware evaluation through its Alpha Lab and Fundamental agents — scoring performance not in aggregate but across expansion, slowdown, contraction, and recovery. Walk-forward backtesting with real friction costs reveals where strategies hold and where they degrade. Cross-regime consistency is a more reliable signal than peak performance in a favorable environment.