By John Lynch, Director of Actuarial and Underwriting Services at Longevity Holdings
In the realm of underwriting accuracy, mortality outcomes versus predictions fluctuate around average life expectancy, influenced by factors like mortality assumption curves and impairment factors. As population diversifies age, it’s crucial to recognize that underwriting status requires continual adaptation. This article explores the necessity of refreshing underwriting standards to align with dynamic mortality trends, emphasizing the risks of outdated assumptions.
For instance, when utilizing the Fasano Mortality Table with a 100% gross mortality rating, the expected age of death distribution resembles a normal distribution model, albeit with a slight right skew.
The median life expectancy for a 75 year-old male non-smoker is 15½ years, meaning the median expected death age is 90½ years old, indicated by the green vertical bar above. For a pool of 1,000 similar insureds, roughly 50% would die at age 90 or younger and 50% would die after age 90. Roughly 5% of the insured pool is expected to live to age 99 and 8 months, highlighted by the red vertical bar. In other words, the 5% tail risk would be roughly 10 years beyond the life expectancy. If a healthy John Doe at age 75 were to get a life expectancy of 15½ years, the 5% mark would be age 99 and 8 months. This highlights a single case of longevity tail risk with projecting out mortality expectations.
Everyone’s health status is in constant flux, and this is especially relevant for older and impaired lives. Life expectancy reports provide a snapshot of health data and project forward anticipated survival curves based on the assessment at that time. The implicit assumption here is that the insured’s underwriting status remains the same. In other words, the mortality multiplier remains constant. This is often not the case. All underwriting assessments are done as of the last medical record date on file. It’s kind of like buying a new car, the price is accurate only until driven off the lot.
A method to conceptualize the dynamic of change in underwriting status over time in aggregate is by comparing it to bond ratings over time. Portfolio managers calculate or reference the rating transition matrix for credit quality movements for assets like bonds. Bonds originally issued at an investment-grade ranking of BBB may get downgraded to a BB, C, or worse, and now have a higher probability of default. Or the reverse can happen, going from a high-yield rating (e.g. BB) back up to an investment grade rating (e.g. BBB, A, etc.).
A similar transition matrix can be created from a large enough dataset of underwriting data with multiple re-underwritings to analyze health statuses over time across various ages and impairments. An attempt to visualize this transition matrix is below. Following the blue data point and line from the “healthy” region at the date of life expectancy estimate to “now” in the “impaired” region is a rudimentary example of a transition. Without updating medical records and performing a new life expectancy report, this information would not be known.
As the medical field continues to evolve and offer new technologies and medicines to help prolong life and cure diseases, we can only expect to see more individuals with impaired ratings to make a recovery and live healthier, longer lives. This is something to be celebrated and is cause for the underwriting field to evolve along with the medical field.
Of course, from the perspective of a life settlement asset holder, such an event is much akin to a bond credit downgrade lowering the market value of the asset. This is longevity risk in action. Therefore, it is important to perform in-depth longevity risk analysis, including Monte-Carlo simulations and sensitivities.
For owners and portfolio managers of life settlement assets, it’s important to refresh life expectancy reports every 2 to 3 years, when possible, to keep up with current underwriting status.
Don’t let stale underwriting and mortality assumptions lead you astray!