Unum 2006 Annual Report Download - page 65

47
Key Assumptions
The calculation of policy and claim reserves involves numerous assumptions, but the primary assumptions used to
calculate reserves are (1) the discount rate, (2) the claim resolution rate, and (3) the claim incidence rate for policy
reserves and IBNR claim reserves. Of these assumptions, our discount rate and claim resolution rate assumptions
have historically had the most significant effects on our level of reserves because many of our product lines provide
benefit payments over an extended period of time.
1. The discount rate, which is used in calculating both policy reserves and incurred and IBNR claim
reserves, is the interest rate that we use to discount future claim payments to determine the present
value. A higher discount rate produces a lower reserve. If the discount rate is higher than our
future investment returns, our invested assets will not earn enough investment income to support
our future claim payments. In this case, the reserves may eventually be insufficient. We set our
assumptions based on our current and expected future investment yield of the assets supporting the
reserves, considering current and expected future market conditions. If the investment yield on
new investments that are purchased is below or above the investment yield of the existing
investment portfolio, it is likely that the discount rate assumption on new claims will be
established below or above the discount rate assumption on existing claims to reflect the new
investment yield.
2. The claim resolution rate, used for both policy reserves and incurred and IBNR claim reserves, is
the probability that a disability claim will close due to recovery or death of the insured. It is
important because it is used to estimate how long benefits will be paid for a claim. Estimated
resolution rates that are set too high will result in reserves that are lower than they need to be to
pay the claim benefits over time. Claim resolution assumptions involve many factors, including
the cause of disability, the policyholder’s age, the type of contractual benefits provided, and the
time since initially becoming disabled. We use our own claim experience to develop our claim
resolution assumptions. These assumptions are established separately for the probability of death
and the probability of recovery from disability. Our studies review actual claim resolution
experience over a number of years, with more weight placed on our experience in the more recent
years. We also consider any expected future changes in claim resolution experience.
3. The incidence rate, used for policy reserves and IBNR claim reserves, is the rate at which new
claims are submitted to us. The incidence rate is affected by many factors including the age of the
insured, the insured’s occupation or industry, the benefit plan design, and certain external factors
such as consumer confidence and levels of unemployment. We establish our incidence
assumption using a historical review of actual incidence results along with an outlook of future
incidence expectations.
Establishing reserve assumptions is complex and involves many factors. Reserves, particularly for policies offering
insurance coverage for long-term disabilities, are dependent on numerous assumptions other than just those
presented in the preceding discussion. The impact of internal and external events, such as changes in claims
management procedures, economic trends such as the rate of unemployment and the level of consumer confidence,
the emergence of new diseases, new trends and developments in medical treatments, and legal trends and legislative
changes, among other factors, will influence claim incidence and resolution rates. Reserve assumptions differ by
product line and by policy type within a product line. Additionally, in any period and over time, our actual
experience may have a positive or negative variance from our long-term assumptions, either singularly or
collectively, and these variances may offset each other. We test the overall adequacy of our reserves using all
assumptions and with a long-term view of our expected experience over the life of a block of business rather than
test just one or a few assumptions independently that may be aberrant over a short period of time. Therefore it is not
possible to bifurcate the assumptions to evaluate the sensitivity of a change in each assumption, but rather in the
aggregate by product line. We have presented in the following section an overview of our trend analysis for key
assumptions and the results of variability in our assumptions, in aggregate, for the reserves which we believe are
reasonably possible to have a material impact on our future financial results if actual claims yield a materially
different amount than what we currently expect and have reserved for, either favorable or unfavorable.