Manager Selection

Section: Manager Selection and Evaluation Estimated study time: 45 minutes Content: Manager selection is one of the most important decisions in institutional investing. Studies consistently show that manager selection (choosing the right active manager) has greater impact on portfolio outcomes than asset allocation for those who delegate to external managers. The manager selection process covers: manager universe screening, quantitative analysis, qualitative due diligence, fee negotiation, and ongoing monitoring. Quantitative evaluation of managers begins with performance statistics but must go beyond raw returns. Key metrics include: Information Ratio (IR = alpha / tracking error), Sharpe Ratio, Sortino Ratio (using downside deviation instead of standard deviation), maximum drawdown, and batting average (percentage of periods outperforming the benchmark). However, performance data must be used cautiously: short track records (less than 3-5 years) have very low statistical significance for distinguishing skill from luck. Even a three-year record with monthly data gives only 36 observations — insufficient to detect small but consistent alpha with statistical confidence. Performance persistence is the central question in manager selection — does past outperformance predict future outperformance? Evidence on persistence is mixed: some studies find short-term (1-year) persistence, particularly in fixed income; long-term persistence is much weaker except at the extremes (top-quartile managers tend to stay above median; bottom-quartile managers tend to stay below median). Candidate explanation for persistence: persistent managers have genuine skill differentiators (process, information, risk management) that are difficult to replicate; non-persistent outperformance reflects luck or exposure to transient factor tilts. Type I and Type II errors in manager selection: Type I error is retaining (or hiring) a manager without skill (false positive — mistakenly believing skill exists). Type II error is firing (or not hiring)…