Role, risk, and residual alpha: A framework for manager research

Given shifting macro and market conditions, many asset owners we’ve spoken to are planning to revisit their strategic asset allocation and consider a range of potential changes, including adding to active equity strategies (amid rising market dispersion), thematic investments (as a complement to growth-oriented investments), and real assets (as an inflation hedge). When the time comes to implement changes like these, an asset owner’s approach to hiring and then assessing managers will, of course, be critical to success. To help, Head of Manager Research Kat Price recently shared her views on manager selection best practices with Head of Multi-Asset Strategy Adam Berger, who added some thoughts of his own.

Adam: As someone who works with multi-manager solutions and helps multi-asset portfolio managers implement their asset allocation views, how would you describe your approach to manager research?

Kat: We use what I refer to as the “three Rs” framework, focusing on 1) the role the manager’s strategy will play in the overall allocation, 2) the risks the manager takes, and 3) the residual alpha potential (i.e., manager skill).

I would highlight a few points about this approach. First, it draws on our Fundamental Factor Team’s factor-based analytical approach; we think factors can be valuable in helping to tease out a manager’s natural style biases and the market environments in which alpha generation may be most or least likely. Second, we find that the significance of “role” is often taken for granted and, in fact, believe it is as important as skill when constructing a multi-manager portfolio.

Finally, we think these dimensions should be assessed on different time horizons. Role and risk are more dynamic, meaning they can be evaluated over a shorter time horizon, while residual alpha potential should rely on a longer-term assessment of a manager’s skill.

Adam: Based on your experience with this framework, can you share some best practices that asset owners might incorporate in their own research approach?

Kat: Sure, I’d offer four ideas:

• Integrate both quantitative and qualitative analysis. These two perspectives can form a virtuous cycle that helps to uncover meaningful insights. To bring this to life, let’s assume a manager has experienced a period of underperformance and there is no single specific factor headwind that explains those results using traditional vendor models (i.e., the quantitative perspective). The asset owner might assume this was a security selection problem. But if the asset owner knows that the manager has a tendency toward higher-quality companies that are also cheaper (i.e., the qualitative perspective), then quantitatively assessing the intersection of quality and value using a proprietary framework could help pinpoint what’s really going on and whether it’s a skill issue or a style issue.
• Have a consistent framework for the evaluation of a manager’s role in the asset allocation. As noted, we think role is as important as skill. While we use a factor framework to identify role (e.g., we find the trend-following factor often aligns with growth-oriented managers in the equity space), there are many ways to approach this process and define roles, including by asset class, sector, region, style, and market capitalization. But ultimately, the point is to have a robust framework and stick to it.
• Assess risk from multiple perspectives. No single tool or metric is going to adequately define risk. This is why in our own process, to offer an example, we use vendor models as well as proprietary tools. We also use what we call reverse stress testing, which involves applying our full library of factor, industry, and historical stress scenarios to a portfolio — to see if any of the results, whether positive or negative, are outside of our expectations.
• Have a disciplined approach for deciding when to move on from a manager. We think it’s important to guard against behavioral biases that can unintentionally influence a decision to change managers. The key is to establish a framework in advance that’s structured and based on a robust set of data. For us, being clear about the role we anticipate for a manager (and how we measure that), the risk profile we are expecting, and the environments we think will be most/least favorable for alpha generation creates a natural framework for deciding when to move on from an allocation.

Adam: How do you think asset owners should think about the right number of managers in a portfolio?

Kat: The unsatisfying answer is that it is whatever number is necessary to achieve the desired risk profile while preserving enough differentiation from the benchmark to deliver alpha in the face of practical constraints, such as capacity. Metrics like active share can be useful when it comes to monitoring the aggregate multi-manager portfolio for differentiation from the benchmark. To maintain that differentiation, it can help to have clearly defined roles for each underlying allocation and low holdings overlap across managers. However, beware that the narrower the role, the greater the number of managers that may be necessary to achieve a diversified profile.

Other considerations for asset owners include operational and business topics, such as key-person risk, operational implications, and capacity. Larger allocation sizes may warrant being broken up across managers to diversify key-person risk and capacity constraints. Conversely, very small allocation sizes may bear operational costs (particularly for separate account investors) and oversight burdens that are disproportionate to the investment impact of the allocation.

Additional perspectives: The links between manager selection and asset allocation

Adam: I’ll close by sharing a few of my own thoughts on manager selection. First, manager selection and strategic asset allocation go hand in hand as drivers of investment success, and there is room for innovation and process improvement in both areas. As Kat notes, one of the keys is building a set of analytics that can be used to mitigate the impact of behavioral biases.

I would also encourage asset owners to strive for balance in how they divide their time between manager selection and strategic asset allocation decisions. In my experience, there’s a tendency to spend much more time on manager selection. Asset owners also seem willing to undo manager hiring decisions more quickly than asset allocation decisions. I think many would be better served by having more patience with manager decisions; although, as Kat notes above, when managers are not living up to their “role” or “risk” expectations, moving more quickly may be prudent. Today, asset owners have almost instantaneous data on managers, but I think that can increase the risk of overreacting to outlier results. Recently, my colleagues and I shared data showing that even the best active managers can suffer extended periods of underperformance — suggesting that asset owners should have a strategy for staying focused on the long term and avoiding costly snap decisions. Read more on the research here.