The art and science of asset allocation and manager selection have become institutionalised in most modern investment programmes, yet most plan sponsors fall short when it comes to translating an allocation strategy and manager roster into a coherent and successful investment programme structure. Risk budgeting is the missing ingredient in this equation, and US sponsors are in the early stages of capitalising on this relatively recent addition to the investment professional’s toolbox.
Risk budgeting should be the final step in a plan sponsor’s overall investment process, but until recently it has been overlooked. To understand why sponsors’ interest in this relatively new area is growing, we need to examine the investment process as depicted in the figure.
The process normally begins by translating one’s goals and circumstances into concrete investment objectives. For a pension fund these objectives typically relate to funded status and contribution levels, while endowed institutions usually focus on spending levels and purchasing power.
Sponsors next devise an asset allocation strategy designed to meet the previously articulated investment objectives. This stage often entails extensive quantitative analysis and results in the specification of a policy portfolio. In its traditional form the policy portfolio expresses the investor’s chosen strategic (ie, long-term target) asset allocation mix as a blend of passive market indices. For a representative US plan sponsor, a policy portfolio might comprise 50% of a US stock index, 20% of a non-US stock index, and 30% of a US fixed income index.
Once an investor has established an asset allocation framework, a selection process ensues that results in a manager roster. Since US sponsors relying on balanced managers (those responsible for tactical asset allocation as well as security selection) represent a dying breed, the line-up normally includes a collection of specialist managers who may span a broad spectrum of traditional active and passive strategies.
At this stage all that remains is to slice up the overall pie into individual pieces for each manager to oversee. Poor implementation in this final step can undermine all of the exhaustive analysis and careful decision-making that came earlier in the process, yet the fact remains that most plan sponsors approach this critical task in somewhat haphazard fashion. Most investors make decisions regarding the relative merits of active and passive management, the suitable number of managers, the appropriate degree of specialisation, etc. In most cases, however, these choices occur without adequate regard for the investment program as a whole and the manner in which the sponsor wishes to pursue its fundamental objectives. This is where risk budgeting comes into play.
The phrase risk budgeting is shorthand for active risk budgeting. To clarify the significance of active risk, let us revisit the very concept of risk within the investment process. Perhaps the most relevant risk for a plan sponsor is the risk of failing to meet one’s investment objectives. The purpose of the asset allocation process is to mitigate this risk through the development of a sound investment policy that embodies a desirable mix of expected return and total return volatility.
Once a sponsor establishes a policy portfolio, the pertinent risk then becomes the failure to achieve investment results consistent with the policy. One can virtually eliminate this risk by replicating the policy portfolio using low-cost passive investing. (This of course assumes that the policy portfolio is limited to traditional marketable securities. The concept is still relevant in cases where such investments account for the bulk of a sponsor’s allocation.)
Because this type of risk arises when investors choose to utilise varying degrees of active management, we use the term active risk. (Other valid synonyms include benchmark risk, manager risk and tracking error.) Active risk is measured using the familiar unit of standard deviation, though it involves the standard deviation of relative returns (actual returns minus benchmark returns) rather than absolute returns.
Just as the asset allocation process requires sponsors to weigh trade-offs between absolute risk and absolute expected return, risk budgeting involves the assessment of different combinations of active risk and expected active return. There exists a continuum between 100% passive alternatives and highly aggressive active strategies, and virtually every plan sponsor’s tolerance for active risk falls between these two extremes. Choosing a point within this spectrum is the first step toward successful implementation.
The second and more complicated step is identifying the ideal allocation to each manager that will maximise the total portfolio’s expected active return subject to the specified active risk limitation. To do so one must specify levels of expected active return and active risk for each individual manager being considered and conduct a form of mean-variance optimisation analysis. If performed with proper rigour, the result of this process will be an investment programme structure consisting of managers and allocations that properly aligns the investor’s objectives, policy, and chosen service providers.
Several noteworthy trends among US plan sponsors have important implications for risk budgeting. Most significant perhaps is the growing acceptance that active risk matters. It is becoming widely acknowledged that active risk can and should be measured, controlled, and allocated with the same level of rigour applied to other aspects of the investment process.
A logical product of plan sponsors’ sensitivity to active risk is interest in risk controlled active management, also known as ‘enhanced indexing’. By making many small discrete bets away from a given benchmark, these investment products provide excess return potential (unlike pure index funds designed to eliminate all such potential) at one-third to one-half the active risk levels of traditional portfolios. It is not a coincidence that sponsors’ interest in active risk coincides with risk controlled active management mandates winning a growing share of US institutional assets.
Yet another evolution relates to the breakthrough in separating decisions concerning total return risk from those related to active risk. Within a risk budgeting framework a sponsor can logically allocate very little active risk to US equities (implying more indexing and/or risk controlled active management) while allocating a relatively large amount of active risk to fixed income investments. Pursuing more aggressive value added strategies within a bond portfolio need not alter the absolute volatility of the asset class, and this understanding has led many US plan sponsors to embrace more innovative approaches to managing a ‘conservative’ asset class.
By decade’s end, we should expect to see risk budgeting acknowledged as a critical step in the investment process, every bit as important and powerful as asset allocation and manager selection. Few American plan sponsors are there today, but the concept is spreading, as are tools and the knowledge to make it work. If the potential is realised, risk budgeting will prove to be a powerful extension of modern portfolio theory that will help institutional investors around the globe maximise their success.
David Brief is senior consultant and director of research at Capital Resource Advisors in Chicago