Defining Quality: What Makes a Good Manager?
When undertaking due diligence on active managers being considered for hire – whether in equities, fixed income or alternatives – it’s important to have a clear objective in mind. Asset owners should ask themselves: what is the motivation for (potentially) using active management for this mandate? What would be the attributes of a high-quality candidate, as compared to a middling one? On what grounds will we differentiate between funds which look very similar in most areas?
Defining what a high-quality manager looks like is case-specific, and not as obvious as it might seem. “Simple: we’re looking for whoever we conclude will be able to generate the most alpha versus the benchmark,” some fund selectors might say. Putting aside the fact that even this straightforward statement can have multiple for interpretations – for example, is “alpha” risk-adjusted, and if so how exactly? – purely maximizing expected alpha rarely captures the entire goal.
Suppose there are two managers. The diligence process concludes that Manager A is expected to generate 7% alpha over a 3-year period, with a standard deviation of 4%. Manager B is expected to generate 5%, with a standard deviation of 1.5%.
Which should be selected? It depends on the specific needs and risk-aversion of the investor: how is this allocation supposed to contribute to the portfolio’s return target? What are the costs of falling short, versus the benefits of doing better? In some situations, Manager B might be viewed as higher quality, despite having lower expected returns.
The Empirically Compass, our framework for evaluating and comparing active strategies, specifies seven dimensions of quality. Taken together, these 7 dimensions can capture the relevant attributes needed to make investment decisions in nearly all manager searches. While the ingredients of quality don’t change, their weighting varies across asset owners, and sometimes across individual searches.
A well-structured diligence process thus requires the following steps:
- Determine how much importance should be placed on each dimension of quality.
- Obtain accurate, forward-looking measures of each dimension for each strategy under consideration.
- Combine these measurements measure using weights which result in a summary measure of quality that reflects the objectives and priorities defined in step 1.
In this framework, the summary measure provides an ordinal ranking of managers which already incorporates all tradeoffs relevant to the selection decision. The dimensions of quality which follow are thus the fundamental building blocks of a scorecard for manager evaluation.
We identified these specific dimensions as fundamental building blocks because rational investors will universally agree as to their desirability. For each dimension, all else equal, more is better than less; where investors will differ is in how they trade off between the dimensions.
Skill is usually the most important component of active manager quality, and represents the investment product’s forecasted ability to outperform an appropriate benchmark over time. To isolate skill, the benchmark must capture the manager’s actual opportunity set and adjust for risk-taking.
The treatment of a strategy’s factor risks – such as beta or style tilts – is one of the most challenging components of skill evaluation, because with the addition of enough factors, the alpha of any performance track record can be brought to zero. The analysis must make a judgment as to which risk factors the manager should receive “credit” for in the form of alpha, and which should be treated as betas that are not evidence of skill. At Empirically, we’ve developed an innovative performance evaluation methodology to do just that.
Consistency measures the reliability of the alpha generation process: are excess returns harvested at a constant rate, or in bursts? Are there runs of strong and weak relative performance, or does the process have a constant mean?
Consistency is valuable because it means fewer disappointing quarters, years and even multi-year periods. Consistency also facilitates more precise and rapid estimation of skill. However, the nature of some investment strategies requires volatility to maximize returns over time. Therefore, Portfolio Managers who are overly focused on tuning their investment process to maximize consistency can end up paying the price in performance.
Tail risk, a related but distinct concept to consistency, means susceptibility to large relative drawdowns. It describes how left-skewed the relative return distribution is.
There is no clear relationship between tail risk and skill; a skilled manager may or may not pursue strategies which have tail risk. While tail risk is to be minimized and avoided if possible, some strategies may necessitate taking tail risk in exchange for high expected full-cycle returns.
Likewise, low-consistency strategies can also have either high or low tail risk, and vice versa. Certain investment strategies – such as those which resemble option selling – can have high consistency for long periods but then experience devastating tail risk. Managers are not always forthcoming about tail risk, so in due diligence – especially hedge fund due diligence – it is important scrutinize strategies which appear to have high consistency to see whether they also possess tail risk which has not yet materialized.
Adaptability considers a strategy’s ability to succeed in diverse market environments, and to alter its process in response to changing market conditions. In a fast-changing world and in highly competitive markets, “one-trick ponies” are ultimately bound to disappoint.
The highest-quality investment strategies are able to add value in a range of economic environment and market conditions, by rapidly adapting or ideally anticipating shifts and responding appropriately. Given that the environment is continually changing in unpredictable ways, the most valuable managers are those whose performance is not contingent upon a narrow set of conditions.
True “alpha” is vanishingly rare; most excess return streams versus common benchmarks (such as a market index) exhibit high correlation to other peers or to common factors, such as momentum or value. Non-correlation of alpha generation process is thus a prized attribute, since – in addition to suggesting true skill, assuming the alpha is positive – it provides a valuable diversification benefit in the asset owner’s broader portfolio.
At the opposite extreme, managers whose excess returns can be easily replicated using certain factors or ETFs are likely a good candidate for replacement by either a lower-cost alternative, or another active strategy offering higher differentiation.
Non-financial attributes encompass a strategy’s conformity to self-proclaimed or investor-targeted non-financial practices or outcomes. This flexible category allows the incorporation of additional objectives – such as manager diversity, Sharia compliance, ESG performance, and impact investing – which are being prioritized in addition to financial performance.
This dimension may be omitted for purely return-maximizing investors; however, an increasing proportion of asset owners are beginning to incorporate additional considerations when selecting external managers. We believe it is essential to incorporate these considerations alongside financial dimensions in an integrated decision making framework, as opposed to applying them as an overlay either at the front end as a “tie-breaker” in a final selection, as doing so allows proper weighing of the relative benefits of the financial and non-financial objectives.
The timing dimension captures the favorability of the present point in time to invest in or divest from a given fund. This variable recognizes that – depending upon their consistency and adaptability – most strategies generate uneven relative return patterns, and can exhibit autocorrelation and mean reversion. As such, we have developed time-series models that can be used to form a dynamic assessment as to the opportuneness of a buy or sell decision.
Given that two investors in an identical strategy can realize dramatically different returns due to differences in their entry and exit points, specific attention to timing as a key component of Empirically’s strategy evaluation process. Proper timing – or even simply avoiding particularly inopportune entry and exit points – can add as much value as a 2-standard-deviation increase in skill under certain circumstances, according to our research.
Putting It All Together
These dimensions form the ingredients of a high-quality manager. Robust measures must be constructed to capture each of them, and then they must be combined into a decision model in a way that reflects each measure’s relative contribution to the overall objectives. Once achieved, this process forms a repeatable and transparent decision making process that will yield the best possible results from a delegated manager program.
Ready to get started?
Schedule a Demo to learn how Empirically can help you accurately measure each of these dimensions, and then rapidly implement a systematic decision process for manager evaluation.
Author Information: Jordan Boslego is a Partner at Empirically.