Anticipating the benefits: About $80 billion is allocated to ordinary life insurance purchases each year, suggesting the paramount importance of understanding differences in policy yields between whole fife and universal life. (Life/Health).

Distinctions between insurance policy types engender different expectations of policy performance. Take, for example, participating whole life and universal life policies, which provide different packages of options. Whole life contracts generally have fixed, level premiums and level death benefits, while universal life policies have flexible premiums, adjustable death benefits, and are unbundled with respect to the various policy charges and interest credits--thus, universal life is said to be "transparent." Participating whole life policies pay dividends.

For several reasons, participating whole life insurance, in theory, might be expected to provide higher cash values than universal life. Participating whole life provides less flexibility than universal life with respect to premiums, partial withdrawals and coverage amounts. Largely because of the lesser degree of flexibility provided by participating whole life, it is less costly to administer than universal life.

The foregoing reasons suggest that participating whole life might provide higher cash values than universal life to compensate for the reduced policy flexibility and because of lower administration costs. In addition, participating whole life might provide higher cash values than universal life to attract buyers who hope to be compensated for the lesser degree of price transparency it provides. On the other hand, universal life might be perceived as "riskier" and, therefore, need to provide relatively higher cash values than participating whole life to induce consumers to purchase the universal life package of options.

Our analysis indicates that universal life policy yields (rates of return) were significantly higher than participating whole life policy yields for five-year and 10-year holding periods from 1988 to 1998. About $80 billion is allocated to ordinary life insurance purchases each year, suggesting the paramount importance of understanding existing differences in policy yields between whole life and universal life.

Policy Yield Method

The analysis is based on the Life Insurance Policy Yield Method. This method computes the net rate of compound interest that must be earned on an investment fund (hypothetical or real), net of term insurance charges, so that at the end of a given time period the fund equals the policy surrender value. Policy yields in the form of rates of return are relatively easy to explain to consumers, and they hold more intuitive meaning than other forms of cost disclosure.

The sample of policies that was examined consisted of two types of cash-value life insurance: participating whole life (29 policies) and universal life (62 policies). Participating whole life contracts included in the sample are those

policies with data from Best's Policy Report on Whole Life with Dividend Histories (1998) and have additional policy data from Best's Flitcraft Compend (1988). Universal life policies included in the sample are those policies (option A--level face amount) with complete data in Best's Policy Report on Universal Life (1998). All policies have face amounts of $100,000 and are for nonsmokers. Data for all policies are for the period 1988 to 1998. Thus, actual policy yields (as opposed to projected yields) were calculated based on appropriate term insurance assumptions. The insurers included in the sample represent more than one-third of participating whole life and universal life in terms of in-force face amount.

Universal Has Higher Yields

Results indicate that, on average, universal life provided significantly higher policy yields compared with participating whole life both for five-year and 10-year periods. The accompanying charts (page 68) illustrate five-year and 10-year policy yields (minimum, mean and maximum) for the sample of participating whole life and universal life policies. The charts show that the minimum five-year and 10-year yields for universal life policies were lower than the minimum five-year and 10-year yields for participating whole life policies. The charts also show that the maximum five-year and 10-year yields for universal life policies were higher than the maximum five-year and 10-year yields for participating whole life policies.

Various explanations could account for the lower policy yields of participating whole life compared with universal life. Consumers may be willing to pay more (accept lower yields) for participating whole life than for universal life because they place greater value on the package of options provided by participating whole life. But, in one form or another, universal life offers a similar package of options, plus more flexibility and more transparency. Thus, based on the package-of-options perspective, the finding of lower policy yields for participating whole life is paradoxical. It is noteworthy that of the 17 insurers surveyed that offered both participating whole life and universal life, 16 offered a universal life policy that had a higher yield than their participating whole life policy, with an average difference in 10-year policy yields of about 300 basis points.

In the life insurance market, where buyer passivity and differing levels of information abound, the incentives for agents to disseminate information are dampened by reverse-competition effects of commission-based products. But knowledge of the disparity in policy yields across policy type is fundamental for consumers to make an informed choice between participating whole life and universal life policies. Knowledge of policy yields also is paramount to an informed decision between term and cash-value insurance.

Many individuals remain unaware of the wide variation in life insurance policy performance (cost), both within and across policy type. For policyowners with existing coverage, the finding of higher policy yields for universal life over participating whole life does not imply that all participating whole life policyowners should replace their participating whole life policies with universal life policies.

Improved policy persistency generally is in the financial interests of insurers, as high lapse rates contribute to increased insurer expenses and are significantly (negatively) related to policy cash values. Since policy yields tend to increase as the policy stays in force, the study illustrates the benefits of persistency for policyowners. The study's findings reinforce the need for improved information to avoid inappropriate policy replacement. Coupled with Limra data on the incidence of policy lapsation, findings suggest that replacement activity is too high, especially during the first five years after policy issuance, but also after the first five years, when policy yields generally begin to turn positive.

Understanding and Choice

Results of this study indicate that wide price differences exist within and across alternative forms of life insurance. Results for this time period (1988 to 1998) and for this study's sample indicate that universal life policies, on average, provided higher yields than participating whole life policies.

The findings from this study call to mind the words of Aristotle: "For theories must be abandoned, unless their teachings tally with the indisputable results of observations. "That is, financial-services professionals claiming the unequivocal superiority of participating whole life should understand that, on average, about 300 basis points per year (over 10 years) are sacrificed when choosing participating whole life over universal life. Decisions between participating whole life and universal life bereft of this information are, by definition, uninformed decisions. Such information must be interpreted carefully, since misuse of the study's results could lead to improper policy purchases, replacements and incorrect perceptions of the relative merits of cash value life insurance, all of which are to the detriment of life insurance policyowners and the life insurance industry.

The findings of this study should help the various parties to the insurance transaction better understand the nature of policy performance and its variation within and across policy types. This increased understanding should, in turn, lead to more competition and a more efficient life insurance market, to the benefit of consumers, financial-services professionals, insurers and regulators.

James M. Carson is Midyette Eminent Scholar in risk management and insurance at Florida State University. Mark D. Forster is president of Finsure, Omaha, Neb.

Average Annual Policy Yields, 1988-1993

The sample of policies examined included 29 participating whole life and
62 universal life contracts with face amounts of $100,000 in which the
insured is a nonsmoker.

Five-Year Average, 1988-1993

         Whole Life  Universal Life

Minimum    -24.6%       -100.0%
Mean       -13.2%         -4.5%
Maximum     -3.5%          7.0%

Five-Year Average, 1988-1998

         Whole Life  Universal Life

Minimum     -3.0%         -6.7%
Mean         1.3%          4.1%
Maximum      4.3%          8.0%

Note: Table made from bar graph