Carrier Portfolios

Abstract

In recent years, the demand for passive investment products in hard-to-replicate asset classes, or exotic betas, has grown faster than financial services firms’ capacity to engineer such products. Full replication of exotic betas is usually impossible or too costly. For example, it is particularly difficult to obtain passive exposure to small-cap equity indices or to alternative asset classes, such as private equity with liquid instruments. The carrier portfolio provides a novel response to this outstanding issue. Unlike methods that require a covariance matrix (or factor model) and solve the usual tracking-error quadratic program, the carrier portfolio relies on a linear program that directly takes into account each return observation and minimizes the total absolute exposure of the underlying assets. Hence, the method does not require assumptions about the return data’s statistical distribution, is amenable to very large-scale optimization problems involving thousands of assets, and can handle problems for which the number of assets is greater than the number of return observations. As a result, the carrier portfolio tends to better track various financial benchmarks with fewer investment vehicles and simplifies the process of creating such portfolios across large numbers of potential investments.