Mortality Assumptions: Where Are We Going?
October  2008

In today’s competitive term life insurance market, direct writers are revisiting their list of pricing assumptions to develop more accurate forecasts of future performance. The mortality improvement assumption is one that is under close scrutiny.

Historically, even companies that did not incorporate an improvement assumption benefited from the secular gains over time as actual mortality emerged more favorably than pricing originally assumed. In today’s term market, however,  there is little opportunity to arrive at competitive premiums without some type of mortality improvement built into the pricing.

But there’s a trade-off. While prices are more competitive, there is less margin for error should experience emerge more adversely than assumed ten, 20 or even 30 years later.

Small Change, Big Impacts

Even a small change in the mortality improvement assumption can have a disproportionately large impact on pricing and profitability. In tests we have conducted, a one-percent incremental increase in the annual mortality improvement assumption can reduce the present value of claims from a block of term life business by up to nine percent. Thus, erring on the conservative side (a low or no improvement assumption) locks in a large margin for potentially positive error, but it also represents a price disadvantage.

What’s more, it can introduce adverse selection as higher quality lives gravitate toward the price leaders’ products. An aggressive assumption wagers that the current high level of lifestyle, environmental and health care improvements will continue into the future more or less at the same rate as in the past. But if future experience does not reflect these optimistic expectations, the block of business ultimately may become unprofitable.

Will Mortality Improvement Continue?

The pricing actuary’s views on the mortality improvement assumption depend on answering three main questions, the first being: Is the high rate of mortality improvement experienced in the 20th century likely to persist into the future?

Optimists may cite that the 20th century sustained a roughly 0.5 percent annual decrease in mortality that has extended into the 21st century so far, and episodes of more rapid improvement occurred as particular medical advances and lifestyle changes were introduced to the population. Others may note that the human body systematically fails past a certain age regardless of fitness, and the 20th century may have been an exceptional period of mortality improvement. In other words, mortality assumption improvements were possible because there was so much implicit “room for improvement” through health and medical improvements (see Chart 1).

Chart 1 Squaring of the Survival Curve.PNG 




How Variable are Annual Mortality Improvement Rates?

Question two is: How variable are annual mortality improvement rates around the long-term average? One side points out that most short-term mortality variance can be attributed to mortality’s random nature but that the long-term trend has been very stable.

The other side notes that while some of the more dramatic downward shifts in mortality can be attributed to particular changes in lifestyle (rapid decline of male mortality in the 1970s due to reduced smoking) and medical advances (strong decline of female mortality through to the 1960s due to improvements in obstetrics), it is possible that other lifestyle changes (much increased rate of obesity across all age groups of both sexes) could actually threaten those mortality gains of the past.

How Do Improvements Translate into Insured Population?
The third question is: How do the mortality improvements in the general population translate into mortality improvement for the insured population? For example, any mortality improvements related to improved cancer treatments or methodologies may be blunted by an underwriting process that has traditionally screened against applicants who had exhibited markers or history of particular cancer risk. If a pool has no individuals that have a certain ailment, it cannot experience improvements in mortality based on greater treatment success for that ailment.

However, as an extension of the discussion above, if the cancer incident rate converges to that of the uninsured population over the life of a closed block of insured risks, then a similar percentage mortality improvement may emerge among the insured population. Potentially increased benefits exist because the lives who demonstrate insurability (and the means to purchase insurance) tend to be the lives that seek routine or extraordinary medical care and are in a position to take advantage of medical breakthroughs.

This discussion highlights the importance of understanding the source of mortality improvements in the past and how companies responsibly address assumptions of future improvements in their current pricing. Transamerica Reinsurance continues to research the impact of mortality improvement on pricing by reviewing its own historical experience, modeling future trends and working with clients to assess this important component of modeling term life insurance products. We look forward to discussion about these issues with our clients in the future and welcome any questions you may have.