American Express 2002 Annual Report Download - page 47

I45 AXP IFINANCIAL REVIEW
benefits of reengineering and cost containment initiatives. In addition, 2001 expenses included an unfavorable net DAC adjust-
ment of $39 million. Other operating expenses increased in both years. The 2002 increase reflects the impact of the technology
outsourcing agreement, which resulted in the transfer of costs from human resources expense, a higher minority interest for
premium deposits related to a joint venture with AEB, and a $44 million net increase in DAC expenses related to AEFA’s third
quarter 2002 adjustment discussed below. In 2001, the increase reflects accelerated investing activities for various strategic,
reengineering, technology and product development projects and a higher minority interest related to the premium deposits
joint venture with AEB. In 2001, other operating expenses included an unfavorable DAC adjustment of $28 million.
For annuity and insurance products, the projections underlying the amortization of DAC require the use of certain assump-
tions, including interest margins, mortality rates, persistency rates, maintenance expense levels, and customer asset value
growth rates for variable products. Management routinely monitors a wide variety of trends in the business, including
comparisons of actual and assumed experience. Management reviews and, where appropriate, adjusts its assumptions with
respect to customer asset value growth rates on a quarterly basis. Management monitors other principal DAC assumptions,
such as persistency, mortality rate, interest margin and maintenance expense level assumptions, each quarter. Unless manage-
ment identifies a material deviation over the course of the quarterly monitoring, management reviews and updates these DAC
assumptions annually in the third quarter of each year. When assumptions are changed, the percentage of estimated gross prof-
its or portion of interest margins used to amortize DAC may also change. A change in the required amortization percentage is
applied retrospectively; an increase in amortization percentage will result in an acceleration of DAC amortization while a
decrease in amortization percentage will result in a deceleration of DAC amortization. The impact on results of operations
of changing assumptions with respect to the amortization of DAC can be either positive or negative in any particular period,
and is reflected in the period in which such changes are made. In 2002, excluding the third quarter, the impact of resetting these
assumptions, along with the impact of unfavorable equity market performance, was an acceleration of $22 million pretax of
DAC amortization. Third quarter impacts are described below.
During the third quarter of 2002, AEFA completed a comprehensive review of its DAC related practices. The specific areas
reviewed included costs deferred and DAC amortization periods in addition to customer asset value growth rate assumptions
(which are typically reviewed on a quarterly basis) and other assumptions including mortality rates and product persistency
(which are typically updated on an annual basis in the third quarter). As a result of this review, AEFA took certain actions that
resulted in a net $44 million increase in expenses in the third quarter of 2002.
❙AEFA reset its customer asset value growth rate assumptions for variable annuity and variable life products to anticipate
near-term and long-term growth at an annual rate of 7%. The customer asset value growth rate is the rate at which contract
values are assumed to appreciate in the future. This rate is net of asset fees, and anticipates a blend of equity and fixed income
investments. Prior to resetting these assumptions, AEFA was projecting long-term customer asset value growth at 7.5% and
near-term growth at approximately twice that rate. The impact of resetting these assumptions, along with the impact of
unfavorable third quarter 2002 equity market performance, was an acceleration of $173 million pretax of DAC amortization.
Going forward, AEFA intends to continue to use a mean reversion method as a guideline in setting the near-term customer asset
value growth rate, also referred to as the mean reversion rate. In periods when market performance results in actual contract
value growth at a rate different than that assumed, AEFA will reassess the near-term rate in order to continue to project its best
estimate of long-term growth. For example, if actual contract value growth during a quarter is less than 7% on an annualized
basis, AEFA would increase the mean reversion rate assumed over the near term to the rate needed to achieve the long-term
annualized growth rate of 7% by the end of that period, assuming this long-term view is still appropriate.
❙AEFA revised certain mortality and persistency assumptions for universal and variable universal life insurance products and
fixed and variable annuity products to better reflect actual experience and future expectations. The company completed a
project to update the mortality table used in pricing universal and variable universal life products and in valuing the associ-
ated DAC. The most recently published life insurance industry mortality table was used as a starting point, and was then
modified based on AEFA’s experience. AEFA also observed that recent persistency of its universal life products was consis-
tently better than expected, and determined the trend justified an improvement in assumed persistency rates. Additionally,
AEFA reviewed and updated persistency assumptions for fixed and variable deferred annuity products. AEFA also reviewed