Financial advisers cannot take risks with their portfolios the way hedge fund managers do. It is important for advisers to measure and control the amount of active risk allowed into their portfolios – some call this “active risk budgeting”. Efficient portfolio construction requires thoughtful active risk budgeting.
When a chief investment officer builds her strategic asset allocation, she sets the amount of risk her client is willing to take. Then she chooses the asset class exposures and weights them, using measurements such as standard deviation and beta, to fit that targeted risk.
Similarly, once you know the weight of the asset classes and it is time to fill out manager selection, the manager must set the amount of active risk she is willing to take. Then she chooses the strategies and weights them, using measurements such as tracking error, to minimize that active risk. Some call this a manager fit analysis. The purpose of a manager fit analysis is not to expel all active risk. Just as in optimising the asset classes in strategic asset allocation, the purpose of a manager fit analysis is to gain as much active return (alpha) as possible for every unit of active risk (tracking error) the client takes.
Experience has shown that a combination of active managers with a low correlation of excess returns often produces superior risk-adjusted equity portfolios than a single manager.
Active risk budgeting becomes a particularly interesting exercise when manager selection teams think about blending fundamental and quantitative strategies. Qualitatively they know that fundamental bottom-up stock pickers with concentrated portfolios have a very different investment process than quantitative, factor-based stock pickers. Intuitively, they would expect fundamental and quantitative investors to be good complements to each other. Their intuition would be right.
But many find it prudent to use more than anecdotes about process to blend managers efficiently. Some use a quantitative process as a second check on manager selection blending. In particular, one can look at the correlation of excess returns, which helps measure the diversification between two managers.
Many fund selectors use multiple managers within an asset class, but in our view many may not be combining those managers properly. If they have equal conviction in two funds, many will weight those two managers equally in the mandate. However, there is something wrong with this equation. Allocating sterling equally across two strategies taking different tracking errors results in an unequal distribution of risk in the combination.
For example, one can create two hypothetical return streams. One is a concentrated fundamental stock picker, Manager F, who takes 550 bps of tracking error and lower tracking error quantitative strategy; Manager Q takes 220 bps of tracking error, an equal sterling allocation across both strategies means that, in actuality, Manager F will dominate between 70 per cent and 99 per cent of the combination’s active risk, depending on the correlation of excess returns.
Active risk budgeting gives a better way to build multi-manager portfolios. If we have equal conviction in two managers’ ability to generate alpha with the active risk we give them, we want to give each manager an equal amount of our active risk.
In the example above, suppose Manager F and Manager Q have a very low correlation of excess returns of 0. To make sure each manager is contributing equally to active risk/tracking error, we can weight Manager F at 30 per cent of the portfolio and Manager Q at 70 per cent of the portfolio. The noteworthy result: this blend has a lower tracking error than either Manager F or Manager Q individually.
Even better, once we have reduced our blended tracking error, our information ratio – meaning how efficient we are being with every unit of active risk we take – is also higher than the same metric for either of our two managers individually.
While the example above is for a blend of two funds, the principles involved in evaluating manager fit are as true for a mandate of two funds as for a mandate of four (or more) funds.
At Goldman Sachs Asset Management, we find that running this type of manager fit analysis is most impactful after identifying a short list of funds run by trusted managers whose process is well understood. In other words, we think this analysis serves best not as an initial screen, but as a final selection and sizing tool. Because past performance does not always indicate future success, many investment officers want strong conviction in their managers and their processes before relying on past return streams to make these sorts of judgments about future allocations.
A great benefit of the return of quant is the ability of manager selection teams to add alpha in a complementary way to existing fundamental equity mandates. Some use a manager fit analysis incorporating tracking error, correlation of excess returns, and information ratio to add an objective process to fund selection and weighting methodology.
The goal in doing so is to avoid putting the client at risk for blown expectations, and to be as efficient as possible with the risk the client has entrusted with the manager.
Brendan McCurdy is executive director, portfolio strategy Emea, strategic advisory solutions of Goldman Sachs Asset Management
Key points
It is important for advisers to measure and control the amount of active risk allowed into their portfolios.
Many fund selectors use multiple managers within an asset class.
Active risk budgeting gives us a better way to build multi-manager portfolios.
