Even those who are not sports enthusiasts are likely familiar with the 2002 best seller Moneyball by Michael Lewis.[1] The book explains how Billy Beane, manager of the Oakland Athletics, used new and rigorous analytics to build a highly competitive baseball team despite limited financial resources. When Beane took over the Athletics salary total was half that of their division’s average and one-third of that spent by the New York Yankees. Beane’s innovation was to challenge the conventional thinking regarding the measurement of player skill. Rather than relying on well-established skill metrics, Beane advanced more effective measures of player ability referred to as sabermetrics. The newer metrics enabled Beane and his coaches to see skill that was not being recognized by other teams. This enabled them to pursue highly skilled players that were not in high demand and, therefore, more affordable. In short order the Athletics became the winningest team in their league while continuing to spend far less on their team salaries. The simple act of rethinking how skill should be assessed changed not just the Athletics fortunes and baseball overall but all sports, forever.

The parallels between sports back in 2003 and active equity management today are uncanny. Conventional portfolio analytics have been around for decades. They are widely used not just to assess past performance and risk but also to gauge manager skill. These analytics are excellent in explaining how a fund has performed and identifying the risk/return tradeoffs undertaken in achieving those outcomes. However, these same conventional analytics provide poor measures of manager skill. Their shortfall lies in the data used in their computation. Conventional analytics use as their inputs either a fund’s return series or its daily holdings. These data are themselves fund outcomes. They do not contain the information necessary to recognize and rigorously quantify manager skill.

Fortunately fund manager skill is identifiable. It is found by relating the decisions the manager makes with the outcomes those decisions generate. Doing so connects cause and effect. Such analytics are referred to as decision-based skill metrics.[2] They are the active management equivalent of Beane’s sabermetrics. Decision-based skill analytics provide asset owners and allocators the ability to identify skilled managers. These same metrics also illuminate skill consistency. This being essential for determining likely future outperformance and making effective allocation decisions.  

If decision-based skill analytics are so darn powerful why haven’t they been more widely adopted? In paraphrasing a review of Moneyball by Cass R. Sunstein and Richard H. Thaler [3] the answer is that: The active management industry has evolved into a “bad equilibrium.” Rather than identifying and computing manager skill directly, the industry continues to rely on the weak skill proxies afforded by outcome-based analytics. Change, as Sunstein and Thaler point out, is difficult even when clear evidence for its necessity is present. And it is made all the more difficult when attempting to introduce change into an industry already besieged with myriad challenges.[3] Ironically, the adoption of decision-based skill analytics might very well lessen the industry’s turmoil. For surely a great deal of the unrest across active management derives from the enormous uncertainty regarding who is highly skilled and who isn’t. Thus, rendering fund selection today more a guessing game than an analytic exercise.

Change is difficult. So are the effects of ineffective allocations, poor performance, and a rapidly shrinking number of active equity providers. This sure looks like the right time to change for the better.