As an example, consider the case of a client portfolio with the asset class allocations described in Table 1. When compared to the target portfolio, the client's current portfolio is overweight equities and underweight alternatives.
A naive rebalance might suggest a substantial reduction in the equity holdings, and a substantial increase in the alternatives holdings. However, looking at the underlying factors and corresponding risk contributions in the figure, and rebalancing to those risk contributions allows for smaller changes. Because factors thread across assets one can reach the target risk attribution through smaller moves. This points to another important advantage of rebalancing across risk factors: it reduces the need for large trades. Furthermore, if the risk exposures are aligned between the target and the client portfolio, one might not even need to rebalance, preventing unnecessary trading costs.
Human Plus Machine—Guided Rebalancing
And then there is the human—the experience and common sense.
Think of three principle use cases for optimization. One is to build the target portfolio. A second is to keep the portfolio within an acceptable range of the target. And a third is to make adjustments as the market changes or as views of the market change. All of these are sensitive to the individual client’s requirements and constraints.
We set the target without an optimization machine. If we keep on top of the portfolio, the straying from the target will be marginal. And if we move from asset to factor space, we can look at variations from the target to understand areas of material bias. Moving back to the mathematical world where the optimization tools and computers reside, adding human judgment creates what can be called, in statistical terms, a Bayesian approach. We create a starting point based on our experience and judgment, which in the Bayesian world is called the prior, and then push it on to the computer to make adjustments, resulting in what is called the posterior. Not that any of this is essential to know in a practical sense, but it is comforting to know that bringing in the human element is fair game from a purist’s standpoint.
Now advisors and their clients are thinking of investments in this way. The abstract point of view that is taken when performing traditional optimization of a portfolio as a simple collection of assets is inadequate when confronted with the reality of how client portfolios are actually set up. The portfolio is effectively a set of sub-portfolios, each with a particular mandate, possibly with underlying accounts with differing objectives. Clients might hold legacy positions or place constraints on the buying or selling of certain assets. At a minimum, any portfolio construction or rebalancing exercise should be cognizant of these realities.
The key point is that mean-variance optimization is too blunt a tool, and one that is difficult to customize to address the needs for a heterogeneous set of clients. The tools that advisors use need to be malleable and flexible to account for this heterogeneity. Modern technology, computing power combined with advances in mathematical techniques, can help advisors move from brute force, machine-driven optimization to what we call guided rebalancing. Guided by the human sense of the baseline and acceptable variations, and guided by optimization methods that respect the realities of the market and the needs and objectives of individuals.
Rick Bookstaber is co-founder and head of risk at Fabric RQ. Dhruv Sharma is head of portfolio intelligence at Fabric RQ.