Active equity investors want their capital in the hands of skilled managers.

The reasoning is straightforward: Managers with the greatest levels of skill should do better than their less skilled peers over time, all other things being equal. Identifying who is highly skilled and who is less so, however, remains a difficult if not utterly impossible task for most investors. And make no mistake about it, this problem is equally vexing to large and highly sophisticated asset owners as well as individuals.  The inability to effectively assess skill lies in the analytics commonly used for such investigations. Such conventional analytics provide useful information for sure – just not about skills.

What’s meant by conventional analytics is the myriad of metrics regularly used in support of fund assessments. These analytics include relative return, multi-factor alpha, information ratio, upside/downside capture, active share, attribution, hit rates, batting average, and slugging ratio. These metrics are effective in describing how a fund generated its returns. They indicate whether the fund’s returns were the result of high concentration, market factor cyclicality, wisely overweighting or underweighting sectors, or taking on additional risk (i.e., volatility). What these analytics cannot do is identify or quantify skill.

The reason is elemental. Conventional analytics are computed using a fund’s return series and/or its holdings history as their data. These data are themselves outcomes. They reflect the performance of the fund which constitutes its returns and also substantially determines the size of its holdings over time. These conventionally derived metrics, therefore, are referred to as outcome-based analytics. And while outcomes do reflect the presence or absence of skill they are not themselves measures of skill.

Meaningful measures of skill are found in the relationships between types of manager decisions and the results they generate. Said differently, skill measurement involves capturing cause and effect. A golfer’s skill is not measured by how many games they win. It is assessed by the distance and placement of drives off the tea, the ability to hit shots with irons close to or onto the green, and the accuracy of the putting. As each of these skills improve we’d expect the golfer to achieve stronger results (i.e., win more games).

As mentioned, manager skills are observed by investigating specific decisions or actions taken and the results they generate. Examples of manager skills that can be isolated and calculated include:

• Buying new stocks that more often than not outperform their sectors or benchmarks.
• Selling positions such that the alpha from strong stocks is captured while the drag from underperforming stocks is minimized. 
• Sizing positions so that sufficient capital is invested in the strongest stocks (i.e., allowing them to lift overall fund returns).

Metrics such as these reflect what are known as decision-based analytics. These analytics capture the cause and effect relationships between manager actions and fund returns. They provide unambiguous measures of skill. Using the results of decision-based analytics enables investors to more effectively assess a fund’s desirability and to make better allocation decisions.

Currently there are a number of firms which provide decision-based analytics.¹  The question that the industry needs to address is why aren’t these superior measures of skill being used regularly today? Absent such analytics investors simply cannot make their best decisions, which can lead only to bad outcomes for all market participants.