1. ^aSchool of Business, Virginia Commonwealth University, Snead Hall, 301 West Main Street, Richmond, VA 23284-4000, U.S.A
2. ^bDepartment of Information, Risk, and Operations Management, The University of Texas at Austin, CBA 5.202, Austin, TX 78712-1175, U.S.A.

Abstract

In this paper we examine the impact of asset allocation strategies on the investment performance of life insurers in the U.S. We are especially interested in comparing the effects of active portfolio management to the effects of passive holding strategies. We define three novel quantitative indices of static/dynamic strategies to represent important dimensions of the active/passive spectrum of investment strategies. The indices are computed from portfolio allocations reported in the firms' annual statement data. Using cluster analysis, we partition the population of life insurers into three groups, characterised by generally having static (passive), dynamic (active) and mixed asset allocation strategies, respectively. There are major differences among the three clusters in terms of risk profile, size and other factors. We model investment performance and allocations among the major holdings of bonds, stocks, cash and mortgages explicitly as simultaneously interacting phenomena. In order to isolate the effect of allocation strategies from the confounding effects of other risks, we imbed a spectrum of enterprise risks, treated as exogenous or predetermined, within the model framework of endogenously interacting performance and asset allocation variables. The strategy indicators are also exogenous variables. We find that the most active cluster enjoys the greatest relative performance, even controlling for allocations among asset classes.