Allocation Betas

Abstract

The complexities of standard optimization can obscure the intuitive decision process that should play a major role in asset allocation. The use of allocation alphas and betas—with U.S. equity as the beta source—facilitates an intuitive approach and greatly simplifies the decision process. A portfolio's assets are separated into two groups: “Swing assets” are the traditional liquid asset classes, such as U.S. bonds and equity; the “alpha core” is all other assets, which are subject to more stringent limits. After the nontraditional assets are combined to form an alpha core, the result is a three-part efficient frontier: (1) a cash-to-core segment, (2) a fixed-core segment, and (3) an equity extension. The boundaries lead to a “sweet spot” on the efficient frontier where most U.S. institutional portfolios are clustered.

The market assumptions behind standard asset allocation studies embed a set of expected returns and covariance relationships among the relevant assets. These assumed relationships are often not consistent with either equilibrium conditions or an efficient market view. This article deconstructs the relationships in an illustrative set of market assumptions into “beta” components that are correlated with U.S. equity and “alpha” components that are independent of U.S. equity. The term “allocation beta” is used to underscore that these values are derived from a covariance matrix intended as the starting point for an allocation study.

The selection of U.S. equity as the beta source is motivated by its role as the dominant risk factor in U.S. institutional portfolios. In addition, the beta measure relative to a U.S. equity index is a familiar and intuitive concept (even though it is not typically used in an allocation context).

When this asset-based analysis is applied to representative U.S. institutional portfolios, roughly 90 percent or more of their volatility is explained by their beta sensitivity to U.S. equities. Moreover, for a wide range of U.S. institutional portfolios, the beta values (and the overall volatilities) tend to be surprisingly tightly clustered around a beta of 0.60 and a volatility of 10 percent.

This framework provides a simplified approach to the allocation process. A portfolio's assets can be decomposed into two groups—“swing assets” and an “alpha core.” Swing assets are the traditional liquid assets—U.S. equities, U.S. bonds, and cash—whereas the alpha core consists of all other assets—non-U.S. equity, real estate, hedge funds, private equities, and so on—that are potential alpha sources but are generally subject to relatively tight portfolio constraints. Although the alpha core is often viewed as serving to diversify the volatility of the swing assets, its real benefit tends to be return enhancement.

The allocation process in this approach can then be viewed as a three-step process. First, maximum acceptable limits are determined for each “nontraditional” asset class. Second, these alternative assets are combined into a “subportfolio”—the alpha core. The composition of the alpha core will generally involve both intuitive and qualitative considerations that go well beyond the explicit quantitative characteristics embedded in the return-covariance matrix. The fund will try to include this alpha core within the final portfolio at its maximum allowed percentage. The third step is to adjust the composition of the swing assets to achieve the desired risk level for the overall fund. (In essence, this process “reverses” the usual path, whereby a portfolio of traditional assets is deployed into alternatives on an incremental basis.)

The assumption of a fixed percentage weight devoted to the alpha core leads to a three-part efficient frontier: (1) a cash-to-core segment, (2) a fixed-core segment, and (3) an equity extension to 100 percent equity. The first and third segments are tied to the fixed points of, respectively, 100 percent cash and 100 percent equity.

Not surprisingly, when four illustrative portfolios are examined in depth, most allocations fall—rather tightly—within the fixed-core segment. Long-term funds apparently have an incentive to move beyond the cash-to-core segment and an incentive to not pursue the equity extension. The fixed-core segment thus forms a “sweet spot” on the efficient frontier, a spot where most real-life portfolios are likely to cluster.