Life insurance portfolio management



Journal Title

Journal ISSN

Volume Title



Based on the institutional and economic environment in which life insurance companies operate, this study develops a theoretical portfolio model with sufficient empirical content to yield hypotheses about life insurance portfolio behavior which are readily tested with appropriate econometric techniques. State quantitative and qualitative restrictions on portfolio composition, the accounting procedures promulgated by the National Association of Insurance Commissioners, Federal tax laws, and other taxes and regulations provide the institutional setting for the life insurance investment process. Given these, the solvency of a life insurance company is affected by the riskiness of its investment portfolio, the uncertainty of future operating expenses, and the uncertainty of cash flows between the policyholder and the company. This study focuses on the first of these sources of risk and abstracts away from the other two. A simple chance-constrained model is presented which integrates the basic economic variables relevant to life insurance company portfolio management. In this model, the firm's objective is to maximize its rate of return on its portfolio subject to a probabalistic solvency constraint, legal constraints, a balance sheet constraint, and non-negativity constraints. Given parameters for the model, the optimal portfolio may be specified, and a sensitivity analysis to parameter changes provides the basis for several testable hypotheses. Several significant economic relationships were found using an econometric analysis of a cross-sectional/time series panel of 92 large U.S. life insurance companies over the 1957-71 time period. The amount of surplus and a proxy for the level of financial yields were positively related to investments in equities and negatively related to investments in bonds and mortgages. Company size was positively related to the proportion of assets invested in common stock and bonds and inversely related to the proportional investment in mortgages. Distributed lag models showed the response to changes in independent variables to be fairly rapid, with the demand for bonds and common stock possessing the highest adjustment rates and changes in the amount of surplus causing the most rapid responses. While New York regulated companies generally invest in less common stock, this result is largely explained by the economic differences between New York licensed companies and other companies. Stock companies invested in more conservative portfolios than mutuals, given the independent variables: this result is expected because of the differences in the life insurance products sold by the two types of firms. Considerable unexplained interfirm variation persisted, perhaps because the high substitutability among financial assets makes prediction difficult. Relative yields often did not prove to be significant determinants of portfolio choices. These results may be due to the problems of multi collinearity, autocorrelation, and sample size, but the poor performance of yields may have other bases, as well. In the chance constrained model, the demand for a security is a complex function and, conceptually, the demand for a security may be inversely related to its own yield. The demand for a security will be inelastic with respect to yield changes when upper or lower bound constraints are binding. In addition, the normative significance of relative yields may be questionable if yield changes occur in response to other characteristics of securities such as maturity, liquidity, and risk of default when these other features are not explicitly accounted for in the model. This study of life insurance portfolio behavior deals with, a number of important theoretical and empirical problems. While the empirical results generally conform to expectations derived from an analytical model, much portfolio variation remains unexplained. At the same time, however, much portfolio variation among the large life insurance companies in this study does appear to have a rational economic basis.



Life insurance