Tuesday, March 10, 2015

Benchmarks, active bets, and shrinking beta


Benchmarks are chosen based both on what is acceptable to investors and past precedent. For US equities, it is the Standard and Poor's 500 and for US bonds, it is usually the Barclays Aggregate Index. A good global equity index is the MSCI world index. Small caps may often use the Russell 2000 index. Pension clients may mix and match indices to complement their long-term objectives while managers will usually want a single generic index for peer comparison and marketing. Picking of the benchmark has strong implications on the measurement of alpha and beta for a manager. It also has large impacts on measuring the active bets taken by a manager.

What is missed in this process is the use of the benchmark as an active tool to determine or measure what are the bets taken by the manager. The chosen benchmark makes a difference because it provides a relative scale for the bets taken. A simple benchmark that is easy to beat will provide a false sense of alpha creation.

Tracking error against a benchmark is a key risk measure, but it is not very informative.  The use of active shares provides another dimension on the quality of active management. The active share is the difference in exposures against a benchmark. Think of it as the position level or micro tracking error of a manager bets against a benchmark's allocation. From the exposure differences comes the tracking error return.

Using benchmarks as an active tool still seems to be in its infancy. This is especially the case with hedge funds that do not always want to track the amount of active bets versus passive indices. This is also fairly difficult to measure for managed futures or global macro managers because the active bets have to be measured across multiple benchmarks.

However, there is some clear conclusions we can draw from looking closely at benchmark selection. If you choose a better benchmark, you will shrink your alpha and increase your beta. If the benchmark better mirrors actual behavior, then it will cover all of the asset that could be employed. This is a good strategic investment measure, but may not be good for the manager who wants to be paid for outperformance.

Choose the right benchmark and you may get better long-term performance even if the positive tracking error is less and the active bets are smaller. Bigger active share bets against a poor benchmark may make the manager feel good and may impress clients, but it may be the wrong long-term choice for the investor.

No comments: