Unum 2006 Annual Report Download - page 66

48
Trends in Key Assumptions
Because our actual experience regarding persistency and claim incidence has varied very little from our policy
reserve and IBNR claim reserve assumptions, we have had minimal adjustments to our persistency assumptions and
claim incidence assumptions during 2005 and 2006. Generally, we do not expect our mortality and morbidity claim
incidence trends or our persistency trends to change significantly in the short-term, and to the extent that these trends
do change, we expect those changes to be gradual over a longer period of time.
Actual interest rates and the assumptions we use to discount our reserves have generally trended downward for all
segments and product lines during 2005 and 2006. Reserve discount rate assumptions for new policies and new
claims have been adjusted to reflect our current and expected net investment returns. Changes in our discount rate
assumptions tend to occur gradually over a longer period of time because of the long duration investment portfolio
needed to support the reserves for the majority of our lines of business.
The mortality rate experience and the retirement rate experience for our block of group pension products have both
remained stable and consistent with expectations.
Claim resolution rates have a greater chance of significant variability in a shorter period of time than our other
reserve assumptions. These rates are reviewed on a quarterly basis for the death and recovery components
separately. Claim resolution rates in our Unum US segment group and individual long-term income protection
product lines and our Individual Income Protection – Closed Block segment have over the last several years
exhibited some variability. Relative to the resolution rate we expect to experience over the life of the block of
business, actual quarterly rates during 2004 through 2006 have varied by +/-7 percent in our Unum US group long-
term income protection and between +2 and -4 percent in our Unum US individual income protection – recently
issued line of business and in our Individual Income Protection – Closed Block segment.
Claim resolution rates are very sensitive to operational and environmental changes and can be volatile over short
periods of time. We consider the recent variability in our claim resolution rate experience to be primarily the result
of a temporary reduction in the operating effectiveness of our Unum US and Individual Income Protection – Closed
Block segment claims management performance. It is possible that this variability could continue into 2007, but we
believe that the actual rates in 2007 will more align with our expected rates. Our claim resolution rate assumption
used in determining reserves is our expectation of the resolution rate we will experience over the life of the block of
business and will vary from actual experience in any one period.
We monitor and test our reserves for adequacy relative to all of our assumptions in the aggregate. In our estimation,
scenarios based on reasonably possible variations in each of our reserve assumptions, when modeled together in
aggregate, could produce a potential result, either positive or negative, in our Unum US group long-term income
protection line of business that would change our reserve balance by +/- 2.4 percent. Using our actual claim reserve
balance at December 31, 2006, this variation would have resulted in an approximate change (either positive or
negative) of $200 million to our claim reserves. Using the same sensitivity analysis approach for our Individual
Income Protection – Closed Block segment, the claim reserve balance could potentially vary by +1.9 percent to -2.4
percent of our reported balance, which at December 31, 2006, would have resulted in an approximate reserve change
in the range of $200 million lower claim reserves to $250 million higher claim reserves. The major contributor to
the variance for both the group long-term income protection line of business and the Individual Income Protection –
Closed Block segment is the claim resolution rate. We believe that these ranges provide a reasonable estimate of the
possible changes in reserve balances for those product lines where we believe it is possible that variability in the
assumptions, in the aggregate, could result in a material impact on our reserve levels, but we record our reserves
based on our long-term best estimate. Because these product lines have long-term claim payout periods, there is a
greater potential for significant variability in claim costs, either positive or negative.