Pension funds that set performance-based hurdles to screen manager searches — such as requiring top-quartile results over the prior three years — might be shooting themselves in the foot.
A recent study of 12,500 institutional strategies by Mercer Investment Consulting Inc., New York, concludes that excellent recent performance not only doesn't guarantee future results but "generally leads to underperformance in the subsequent period."
Terry A. Dennison, a Mercer senior consultant in Los Angeles and the study's author, said in a telephone interview that the results raise the tantalizing question of whether past performance is actually a "slightly negative indicator" of future returns.
If so, the widespread use of performance-based screens to identify "hot" managers — understandable in organizations where decision-makers have to defend their choices — might be "just about exactly wrong," the study concluded.
Mr. Dennison looked at institutional offerings of more than 2,300 investment managers globally to see how often strategies that posted above-median or top-quartile results for an initial three-year period repeated that feat in the subsequent three years. In all cases, the results fell below what a random coin toss would have predicted.
For example, while at least 50% of the portfolios that delivered above-median gains for the initial period might have been expected to repeat over the following three years, Mercer's study of managers in 15 separate domestic and international investment styles showed only the small- to midcap core and value segments coming close, at 46.9% and 45.7%, respectively.