Identifying Unrealistic Expectations in Manager Selection

When evaluating an active manager, most gatekeepers and allocators rely on the past as their guide to the future.  That’s most obvious when it comes to performance — managers whose results are below a certain threshold are weeded out by rule or tradition.

Qualitative judgments are backward-looking too.  They are seen through the lens of performance:  it is natural to praise people or process or culture when the numbers provide the validation for those assessments, even if they are based upon very little in the way of in-depth knowledge about the manager attributes in question.

Thoughtfully setting — and stating — expectations is an essential part of the manager selection process, but looking backward can lead you to the wrong conclusions.

(The examples used below reference open-ended investments that can be entered and exited at any time.  Similar issues exist for lockup vehicles but aren’t examined here.)

The question

An effective way of surfacing expectations inherent in a selection process is by posing a simple question to those involved:  “What is the probability that this manager/strategy/vehicle will outperform by an average of fifty basis points a year over the next ten years (and survive)?”

(As a self-test, you could pause here and answer the question in regard to one or more managers with which you are familiar.  That will give you an idea of how you would approach the question and the probability you think is required to warrant a selection.)

There is nothing magical about fifty basis points rather than some other level — or you could use a risk-adjusted measure instead of excess return.  The point is to provide a specific prediction.

Within an organization, having everyone who is part of a selection process state their assessment of the probability surfaces the range of opinions that exists more precisely than usually happens.  Then, a final prediction approved by the decision makers in the group grounds expectations for the future regarding performance.  Implemented across every selection decision over time, the process also yields valuable information about the accuracy of individual and group predictions.

Context

The purchase recommendation from the initiating analyst and the document reporting the final decision should state the probability and context for it, starting with the applicable base rate.  For example, only a small sliver of mutual funds in most categories outperform a passive alternative over a ten-year period, to say nothing about adding a fifty-basis point hurdle on top of that.

Reporting the base rate as a matter of course should force a discussion about luck versus skill and prompt a careful examination of the qualitative attributes of the manager:

Since a common approach is unlikely to produce the desired return, what aspects of the manager’s organization and methods are truly differentiated?

What evidence of those differences can be independently identified (exclusive of the manager’s narrative and historical performance)?

Which differences rise to the level of competitive advantage or “edge”?

Then the question shifts to the future and the crucial element of time, which is why the bottom-line question is worded as it is.  Research shows that a three-to-five year decision window is a poor framework on which to judge a manager, yet surveys and observed experience indicate it is by far the most common approach.  While even a decade isn’t long enough to show statistical significance, it is a more practical time horizon.  Using that as the foundation for predictions can help limit noise-induced transactions triggered by interim performance concerns.

The element of time brings us face to face with issues that must be addressed:

Has the attractive historical performance been the result of valuation changes stemming from the increased popularity of the underlying securities?  If so, will that revert in the coming years?

Why won’t the identified edge(s) be worn away over time?

Why will the “consistent and repeatable process” (at least as advertised) continue to work?  What are the most likely reasons that it would quit working?

What evidence is there that the management organization is making changes that will allow it to succeed in the future, rather than just continuing to do things as it has?

Shifting the orientation to the future moves the burden of analysis from the historical characteristics and performance to the qualitative assessment of the changing investment environment and the organization’s prospects for growth (in capability, not size) and its adaptability.  Those attributes are commonly underweighted in manager selection even though they are key determinants of future success.

Unrealistic expectations

There are a variety of ways that unrealistic expectations can inhibit the stated goals of manager selection.  The most basic comes from harboring a belief in active management without having the ability to deliver on that promise due to a lack of time, talent, or other resources.  (Nobody likes to admit those shortcomings and few keep records of selection success or failure that could provide helpful evidence.)

One pervasive problem is a lack of education (and ongoing reminders) about typical performance patterns.  If you want to take advantage of active management, you can’t obsess about tracking error or ignore the reality that the best managers often underperform for a fairly large percentage of the time and for extended periods.

Too often, impatience is designed into the selection and ongoing monitoring of managers because of a lack of communication about the nature of performance patterns and the inherent variability of results.  Individual clients without an investment background often don’t understand that managers shouldn’t be expected to outperform regularly, so their expectations can be unrealistic.  More surprising perhaps is that the same tendency often holds for more sophisticated parties, including accredited investors, due diligence analysts, investment committees, etc.

In fact, a firm that trumpets its unique access and ability to select the best managers in the world can often be the most impatient, since it has defined its role in a way that is unlikely to be fulfilled.  If a manager that is expected to stay in the top decile as it has in the past doesn’t do so, the story can wear thin quickly, prompting a swap for another “top-decile manager.”

Let’s get back to that question:

“What is the probability that this manager/strategy/vehicle will outperform by an average of fifty basis points a year over the next ten years (and survive)?”

Some real-world responses:

Due diligence analysts often answer the question with a number in the 65-75% probability range.

When asked what their chief investment officer or investment committee would want to hear as a probability rating before approving an allocation, responses can be more than 10% higher.

After stating that a manager was expected to outperform by two hundred basis points, one analyst was asked what the probability of that occurring was.  “Over 80%.”

Each of those examples (and others like them) point out the power of the general form of the question.

It forces you to respond by identifying the attributes that support your prediction.  More importantly, the consistent use of the question in an organization helps those involved create realistic expectations and avoid cycles of disappointment and bad decision making.

 

The Advanced Due Diligence and Manager Selection course dives deeply into these and related issues; see also the other postings in the due diligence category of the archives.  For information on ways to analyze your selection process, please reach out.