Both dividend illustrations on participating policies and the interest credited on “excess” premiums and cash values in current assumption and universal life policies depend on the assumed mortality charges. Lower assumed mortality charges allow an illustration to show favorable results at a lower premium cost. In addition, some companies show illustrations that assume mortality will improve in the future. However, if actual mortality experience is worse than assumed, the long-term cost of the policy will increase either because dividends are lower than projected or because direct charges to cash values are greater than assumed. In either case, net premiums effectively will have to increase from the level projected to maintain benefits as projected.
Care should be exercised in selecting policies with lower than average assumed mortality charges. Is it reasonable to assume a company may be so much more selective in its underwriting that it can maintain lower than average mortality costs even well beyond the ten- to fifteen-year period when the benefit of selective underwriting is usually assumed to completely “wear off”? If a company that has used aggressive mortality assumptions or has projected mortality improvements turns out to be correct, it is also quite likely other companies that have used more conservative assumptions will also experience better than projected mortality. These companies will be able to pass on substantially better results in future years than are currently being illustrated. All else being equal, it may be better to select companies using more conservative mortality assumptions. This may reduce the risk of having to pay additional premiums in the future to maintain the level of benefits when such payments may be most difficult to sustain. The alternative may be a decrease in benefit levels. If mortality experience is better than assumed, few policyowners will complain if their premium costs decline or vanish sooner than projected.
Reproduced with permission. Copyright The National Underwriter Co. Division of ALM